an investigation into the object relational - UWC Theses and ...

AN INVESTIGATION INTO THE OBJECT RELATIONAL
PATTERNS OF VIOLENT MALE JUVENILE OFFENDERS.
Jean-Philippe Arzul
A mini-thesis submitted in partial fulfilment of the requirements for the MPsych
Degree in Clinical Psychology in the Department of Psychology, University of the
Western Cape.
December 2005
Key Words: Violent Crime, Homicide, Rape, South Africa, Male Juvenile Offenders,
Psychopathy, Thematic Apperception Test (TAT), Social Cognitions & Object Relations
Scale (SCORS), Object Relations, Personality.

ACKNOWLEDGEMENTS
I would like to thank the participants for their valuable contributions to this study, as well
as the guidance of my supervisor, Professor Charles Malcolm.

ABSTRACT
Although deficits in object relations patterns have been identified in populations of violent
offenders, few studies have examined the object relations of male juveniles incarcerated for
violent crimes. The present study examined four dimensions of object relations, as
measured by the Thematic Apperception Test (TAT) and Westen's (1985) Social
Cognitions and Object Relations Scale (SCORS), with a sample of eight male juvenile
offenders incarcerated for violent crimes at De Novo and Eureka Youth Care Centres.
These dimensions are complexity of object representations, affect tone of relationship
paradigms, capacity for emotional investment in relationships and understanding of social
causality. Violent male juvenile offenders displayed complex object representations, a
tendency to attribute simple but logical motivations to others, a malevolent object world
and a relative incapacity to invest in others in a non-need-gratifying way. The above levels
of object relations are illustrated with TAT excerpts and comparisons are explored between
the results and literature findings. Finally, recommendations and therapeutic implications
are outlined.

TABLE OF CONTENTS
PREFACE
ACKNOWLEDGEMENTS
ABSTRACT
TABLE OF CONTENTS
CHAPTER 1: INTRODUCTION...................................................................1
1.1. JUVENILE VIOLENCE..............................................................................................1
1.2. JUVENILE SEXUAL VIOLENCE.............................................................................2
1.3. ASSAULTS ON TEACHERS & LEARNERS...........................................................3
1.4. CRIME STATISTICS..................................................................................................4
1.5. JUVENILES IN PRISON.............................................................................................6
1.6. CONCLUSION..............................................................................................................8
CHAPTER 2: LITERATURE REVIEW.....................................................10
2.1.THE DEVELOPMENTAL CHALLENGES OF ADOLESCENCE.......................10
2.2. DILEMMAS IN DIAGNOSIS OF JUVENILE PSYCHOPATHOLOGY............11
2.3. RESEARCH INTO ADOLESCENT & ADULT PSYCHOPATHOLOGY..........13
2.3.1. Fantasy & Aggression...............................................................................................13
2.3.2. Fantasy, Aggression & Punishment.........................................................................14
2.3.3. Rorschach Studies.....................................................................................................14
2.3.4. Conduct Disorder, Psychopathy & Personality Disorder......................................16
2.3.5. Hostile Attributional Bias.........................................................................................18
2.3.6. The Thematic Apperception Test (TAT) & The Social Cognitions & Object
Relations Scale (SCORS)....................................................................................................19

2.3.6.1. Complexity of Representations................................................................................19
2.3.6.2. Affect Tone..............................................................................................................20
2.3.6.3. Capacity for Emotional Investment in Relationships...............................................21
2.3.6.4. Understanding of Social Causality...........................................................................21
2.3.6.5. Research with the TAT & SCORS...........................................................................22
2.3.6.5.1. Adolescent Object Relations.................................................................................22
2.3.6.5.2. Adolescent Borderline Object Relations...............................................................22
2.3.6.5.3. Adult Borderline Object Relations........................................................................23
2.3.6.5.4. Developmental History & Object Relations..........................................................24
2.3.6.5.5. Object Relations of a Serial Sexual Homicide Offender.......................................25
2.4. CONCLUSION............................................................................................................26
CHAPTER 3: METHODOLOGY................................................................27
3.1. STUDY DESIGN.........................................................................................................27
3.2. PROJECTIVE TESTING..........................................................................................27
3.3. METHOD.....................................................................................................................29
3.3.1. Characteristics of the Sample...................................................................................29
3.3.2. Recruitment & Procedure........................................................................................29
3.3.3. Research Tool............................................................................................................31
3.3.4. Data Analysis.............................................................................................................32
3.4.LIMITATIONS & DELIMITATIONS......................................................................33
3.5. RESEARCH HYPOTHESIS......................................................................................34
CHAPTER 4: RESULTS & DISCUSSION.................................................35
4.1.DESCRIPTIVE STATISTICS....................................................................................35
4.2. MEAN SCORES ACROSS ALL DIMENSIONS....................................................36

4.2.1. Complexity of Representations: Mean Scores........................................................36
4.2.1.1. TAT excerpt: Simple traits – Just Bad.....................................................................37
4.2.1.1.1. Commentary on TAT excerpt: Simple traits – Just Bad.......................................38
4.2.1.2. TAT excerpt: Antithetical traits – The Good & the Bad..........................................38
4.2.1.2.1. Commentary on TAT excerpt: Antithetical traits – The Good & the Bad............39
4.2.2. Complexity of representations: Individual Mean Scores......................................40
4.2.2.1. TAT excerpt: Complex representations...................................................................41
4.2.2.1.1 Commentary on TAT excerpt: Complex representations.......................................42
4.2.2.2. TAT excerpt: Pathological representations..............................................................42
4.2.2.2.1 Commentary on TAT excerpt: Pathological representations.................................43
4.2.3. Affect Tone: Mean Scores.........................................................................................43
4.2.3.1. TAT excerpt: A Violent & Aggressive World.........................................................44
4.2.3.1.1. Commentary on TAT excerpt: A Violent & Aggressive World...........................45
4.2.3.2. TAT excerpt: A Capricious & Painful World..........................................................46
4.2.3.2.1. Commentary on TAT excerpt: A Capricious & Painful World............................47
4.2.4. Affect Tone: Individual Mean Scores......................................................................47
4.2.4.1. TAT excerpt: Variations in Affect Tone..................................................................48
4.2.4.1.1. Commentary on TAT excerpt: Variations in Affect Tone....................................49
4.2.5. Capacity for Emotional Investment: Mean Scores................................................49
4.2.5.1. TAT excerpt: A Need Gratifying Orientation..........................................................50
4.2.5.1.1. Commentary on TAT excerpt: A Need Gratifying Orientation............................51
4.2.5.2. TAT excerpt: Violence with No Remorse................................................................52
4.2.5.2.1. Commentary on TAT excerpt: Violence with No Remorse..................................52
4.2.6. Capacity for Emotional Investment: Individual Mean Scores..............................53
4.2.7. Understanding of Social Causality: Mean Scores..................................................54

4.2.7.1. TAT excerpt: Simple Internal Motivation................................................................54
4.2.7.1.1. Commentary on TAT excerpt: Simple Internal Motivation..................................55
4.2.8. Understanding of Social Causality: Individual Mean Scores...............................55
4.2.8.1. TAT excerpt: Complex Internal Causes Interspersed with Minor Logic Errors......56
4.2.8.1.1. Commentary on TAT excerpt: Complex Internal Causes Interspersed with Minor
Logic Errors...........................................................................................................................58
4.3.CONCLUSION.............................................................................................................59
CHAPTER 5: RECOMMENDATIONS & CONCLUSION......................61
5.1.Structural recommendations.......................................................................................61
5.2.Diagnostic recommendations......................................................................................61
5.3.Treatment implications................................................................................................62
5.4.Research implications..................................................................................................64
5.5.Conclusion.....................................................................................................................65
REFERENCES...............................................................................................67
APPENDICES.................................................................................................75
APPENDIX A......................................................................................................................75

1
CHAPTER 1: INTRODUCTION
1.1. JUVENILE VIOLENCE
Limited scientific information exists on the prevalance and patterns of juvenile violence in
South Africa. For instance, despite the Department of Correctional Services' (DCS)
acknowledgment that the prison population is "becoming increasingly younger, with about
42 percent of the offender population under the age of 25", the crime categories of juvenile
offenders are not noted (DCS, 2005: 24). Instead, crime categories for the overall prison
population, including females, are presented. Much of the information pertaining to juvenile
crime presents itself in the parliamentary forum, or at public meetings addressed by relevant
ministers. These forms of information make their way into the public domain primarily via
the written media, necessitating a reliance on this medium to assess current trends of
juvenile violence.
A number of horrifying homicides and sexual offences committed by South African
juveniles in recent years have begun to draw increasing attention to local juvenile crime.
The gang rape and repeated stabbing (53 times) of Valencia Farmer in Eerste River in 1999
resulted in a prison sentence of 23 years for a 17 year old male; the longest sentence to be
imposed on a juvenile (Barday, 2001). The fatal stabbing of John Rubithon, a well known
political activist and photographer, in his Woodstock home by a 14 year old male burglar in
2000, also received a fair amount of media attention (Article 40, 2000). Some of the
incidents gracing our front pages during the course of 2004 included the sexual molestation
and drowning of a 12 year old boy by his male juvenile friends in Belhar, the assault and

2
drowning of an 8 year old boy by three boys of 7, 8 and 9 years of age, the kicking to death
of a homeless man by male children in Tygerberg, and an adolescent's murder of his mother
in Pretoria (Die Burger: 6 May, 2004; Die Burger: 10 July, 2004).
Further gruesome acts of violence committed by male juveniles during the course of 2005
included the well publicized murder of a 6 month old baby, the stabbing of a 17 year old
male 30 times, and the serious injury of a 10 year old boy following the stoning of a moving
vehicle (Die Burger: 11 July, 2005; Die Burger: 20 September 2005). One of the more
publicized cases involved the murder of an unknown man by the "Waterkloof-four"; all 16
year old males at the time of the murder (Die Burger: 15 June, 2005). The fatal shooting by
a 15 year old boy of his father, the fatal stabbing of a juvenile by a 13 year old boy, and the
arrest of two 17 year old boys for the murder of a 9 year old child caught in cross-fire
between two groups of youths have also attracted much publicity (Die Burger: 11 April,
2005; Die Burger: 5 September, 2005).
1.2. JUVENILE SEXUAL VIOLENCE
Statistics reveal shocking increases in sexual violence on schoolgrounds in recent years; 20
charges were reported in 2001 contrasted against 111 between January and October 2004
(Die Burger: 7 October, 2004). A report by Chisholm & September (2005) revealed that
24% of sexual offenders, who are primarily male, were between 7 and 14 years of age and
steadily increasing. Evidence of these statistics revealed themselves in the recent rape of a 7
year old girl by a school-worker and two boys of 8 and 9 years while at school (Die Burger:
13 April, 2005). Even more worrying is the single rehabilitation centre for young sex-

3
offenders within the Western Cape Province, as well as the practice of generally not
prosecuting or rehabilitating sexual offenders under the age of 14 due to the young age of
the accused and limited medical evidence in substantiation of the purported crime (e.g. the
lack of semen in the victim's vagina) (Wood, 1998).
1.3. ASSAULTS ON TEACHERS & LEARNERS
Physical assaults on teachers by learners appear to be increasing in South Africa, with a
quarter of teachers in Free-State schools reporting assaults on themselves (De Wet, 2005).
The Western Cape schoolgrounds paint a gloomy picture after ranking highest in school
levels of violence. Incidents involving weapons amounted to 22%, assaults to 18%, and
fights with weapons to 14.4% (HSRC, 2005). Statistics such as these go some way in
explaining a Western Cape teacher's recent court action after an assault by a learner in her
classroom (Die Burger: 15 June, 2005). Another report detailed the physical assault on a
pregnant teacher by a pupil in Belhar, while two teachers were injured in protest action by
learners in Fishhoek (Die Burger: 18 March, 2005; Die Burger: 28 February, 2005).
Assaults on fellow learners have increased dramatically in recent years according to the
Minister of Education: 12 charges were laid in 2001 contrasted against 99 between January
and October 2004 (Die Burger: 7 October, 2004). Juvenile male-learner on juvenile malelearner
school violence during the course of the year saw fatal stabbings in Athlone and
Khayelitsha, the fatal beating of a 15 year old by a classmate, the near fatal stabbing of a
learner in Langa, and the stabbing of a 17 year old learner by five youths between 16 and 18
years of age while on his way home from school in Mannenberg (Die Burger: 24 August,

4
2004; Cape Argus: 18 January, 2005; Die Burger: 6 May, 2005; Die Burger: 22 July, 2005).
A friend of the deceased was also stabbed a number of times in the back, precipitating
discussions with the Department of Community Safety regarding the deployment of security
personnel at 35 schools in high-risk areas (Die Burger: 9 August, 2005). Statistics reveal
that 90% of all South Africans between the ages of 11 and 14 have witnessed violence,
while 47% have been assaulted. Even more disturbing are the 14% of South Africans
between the ages of 16 and 25 who have witnessed a murder, totalling some 4 million
persons (Mistry, 2004).
1.4. CRIME STATISTICS
The frequent under-reporting of crime to police services results in difficulty deciphering
crime statistics. Homicide statistics are however one of the more accurate measures of
crime levels as murder is the most likely crime to be reported to the police (Thomson,
2004). The 2002/2003 South African Police Services (SAPS) Annual Report reassuringly
cited a continuing decrease in overall murder rates since 1994, although the 2004 average
national figure of 43 murders per 100 000 citizens unfortunately compares with the murder
rate of the most dangerous urban area in the United States – Washington D.C.. South
Africa's incidence of murder is undoubtedly still serious when one compares our national
homicide rate of 43 per 100 000 to Albania's murder rate of 28 per 100 000, Russia's of 21
per 100 000, Brazil's of 19, the U.S.A. of 5.6 and most of Europe at 4 (Thomson, 2004).
Murder remains the second highest cause of death after Aids in South Africa (Rustomjee,
2005).

5
More worrying is the the Western Cape's status as the only province with an increase of
34% in murder levels since 1994/1995. The Western Cape boasted South Africa's highest
murder rate of 85 murders per 100 000 citizens in 2002/2003, exceeding Gauteng's average
of 59 murders per 100 000, and far outdoing Colombia, the “murder capital of the world”,
whose homicide rate averaged at 66 murders per 100 000 for the same period (Legett,
2003: 62; Legett, 2004a). This figure has fortunately decreased by 25% during 2004 to 60
murders per 100 000, reportedly due to more visible policing and specific operations,
although the figure still remains alarming (Du Plessis & Louw, 2005). In addition to the
highest murder rate in the country, the Western Cape is experiencing the worst overall
crime problem nationally (Legett, 2004a). Khayelitsha seems to have been a hot spot for
some time now, as evidenced in the escalation of robbery with aggravating circumstances
by 306% between 1998 and 2001 (Die Burger: 12 April 2005). The Minister of Safety and
Security, Charles Nqakula, recently declared Khayelitsha as having the highest levels of
murder, attempted murder, rape, and attempted rape in South Africa. The Minister went on
to attribute much of the criminal activity to ten groups of gangs with 60% of members
under eighteen years of age and many under twelve years of age (Die Burger: 26 May,
2004). Similar concerns have been identified in Mannenberg, with two-thirds of the male
population between the ages of 10 and 30 involved in gang activity (Legett, 2004b). Legett
(2003: 5) explains the need for “intense attention” in the Cape if the situation is not to
“swing entirely out of control”.

6
1.5. JUVENILES IN PRISON
Statistics provided by the Department of Correctional Services (DCS) revealed an increase
of 159% in the number of children serving prison sentences between January 1995 and July
2000. This contrasts to an increase of only 33% for 18-20 year olds, 25% for 20-25 year
olds, and 16.25% for prisoners older than 25 (Muntingh, 2001). The Annual Reports of the
DCS between March 2001 and January 2004 do however indicate a more or less stable
juvenile prison population (sentenced and unsentenced) of approximately 4000 inmates
(DCS, 2001; DCS, 2002; DCS, 2003; DCS 2004). Male juveniles awaiting trial as at 31
March 2004 amounted to 2 166, while the female juvenile population totalled 66. The
sentenced male juvenile population stood at 1 890, while the female equivalent came in at
36 for the same period. The overall juvenile prison population, both sentenced and
unsentenced, thus stood at 4158, with male juveniles significantly outweighing their female
counterparts (DCS, 2003; DCS, 2004). Crime categories for the juvenile population do not
exist, although the overall escalation of sentenced offenders for aggressive crimes over the
last three years would suggest a similar increase in the juvenile population (DCS, 2003;
DCS, 2004; DCS, 2005). Proportionately, violent crimes account for 33% of all crimes in
South Africa (Du Plessis & Louw, 2005).
The apparent stabilization of the juvenile prison population may in part be attributable to
the increasing use of diversion programmes to reduce the number of children in prison,
stem the tide of prison overpopulation, and rehabilitate young offenders (Muntingh, 2003).
During the course of 2003 the National Institute of Crime Prevention and Rehabilitation of
Offenders (NICRO) handled the diversion of 17 724 offenders, while 6000 offenders were

7
managed by other institutions. Overall, 50 000 offenders have been diverted to date
(Sekhonyane, 2004). Further reductions in the juvenile prison population took place during
2005, with 1000 children being released in an amnesty bid that aimed to stem the tide of 4
out of every 1000 South Africans living behind prison bars (Fagan, 2004). In addition, the
1, 389 children awaiting trial since January 2005 had been reduced to 885 by June of the
same year (Die Burger, 14 April, 2005; Die Burger: 2 July, 2005). It nevertheless remains
unclear whether successful rehabilitation occurred prior to these releases. This does
however seem highly unlikely given the recent court action by the Prison Care and Support
Network regarding the "unlawful" and "unconstitutional" conditions in prisons that include
ritual anal rape, high HIV-Aids rates, overpopulation by 300% and corrupt officials (Die
Burger: 16 September, 2005).
Most sentenced and awaiting trial juveniles find themselves located in sections of adult
prisons as only four Youth Care Facilities exist nationally. Youth Care Centres (previously
known as “Reform Schools”) are unevenly spread nationally, with one facility in
Mpumalanga and three in the Western Cape. Sentenced male children in the Western Cape
may find themselves placed at Faure, Eureka or De Novo Youth Care Centres. Each of
these Facilities caters for 120 juveniles, explaining the large number of children held in
adult prisons (Blose, undated).

8
1.6. CONCLUSION
The high crime levels in the Western Cape are baffling criminologists who link crime to
deprivation, as the province enjoys the lowest unemployment levels in the country, has a
more equitable distribution of income than any other province, and the lowest percentage of
informal housing and households without water (Legett, 2004a). Increased migration and
urbanization, substance abuse, overpopulation, taxi-violence, and gangsterism partially
explain spiralling trends in crime, although these sociological phenomena do not adequately
address why other areas with analogous social experiences are not similarly affected
(Legett, 2004a; Legett, 2004b; Thomson, 2004).
While various forms of social deprivation most certainly have a contributing role in the
development of juvenile crime, interpersonal expectations shaped by past experiences have
been shown to influence pervasively the current processing of social information (Westen,
1985). The formation of such interpersonal expectations of self and other, known as object
relations, can be understood as “momentarily activated schemata that reflect current
situational cues, ... developmental history and characteristic ways of processing social
information” (Westen et al., 1991a: 407). These internal working representations do include
conscious components, although they are understood to be “largely unconsciously activated
and inaccessible to self-report” (Westen et al., 1991a: 407). Accordingly, the diminishment
of social deprivations may not result in a decrease of antisocial behaviour if accompanying
social representations remain malevolently skewed. And while more research into current
social circumstances contributing towards increased levels of juvenile crime certainly is
required, a dearth of material exists regarding the personality characteristics contributing

9
towards adolescent distress and dysfunction (Westen & Chang, 2000). Given the increase in
juvenile offending in South Africa, an examination of the personality attributes of male
juveniles incarcerated for violent crimes is sorely needed to provide insight into the
psychological dimensions of the spiralling levels of crime in South Africa, and particularly
the Western Cape.

10
CHAPTER 2: LITERATURE REVIEW
2.1. THE DEVELOPMENTAL CHALLENGES OF ADOLESCENCE
Adolescence is often understood as a distinct and homogeneous developmental phase
occurring between childhood and adulthood, with frustrated attempts at defining a sense of
identity, self-concept and self-esteem often leading to emotional and behavioural
difficulties, including Oppositional Defiant Disorder, Conduct Disorder, and juvenile
delinquency (Osborne, 1996). This “at-risk” phase is characterised by a multitude of
changes, including pubertal development, cognitive development, affective changes, school
transitions, changing social roles, and the emergence of sexuality (Gacono & Meloy, 1994:
45; Westen & Chang, 2000). In addition, a marked shift of focus from concrete and external
attributes to internal thoughts, feelings and personality dispositions has been noted (Westen,
1985). The complexity of the adolescent phase thus makes the understanding of
accompanying psychopathology all the more complex.
The American Psychiatric Associations' (APA) curtailment of diagnosing personality
pathology before eighteen years of age further complicates the clinician's understanding and
treatment of violent juvenile offenders (American Psychiatric Association, 2000). While
attempting to protect the premature closure on personality development, this seemingly
arbitrary distinction stands in stark contrast to a number of studies demonstrating the
possibility of Personality Disorder diagnosis in adolescence and its continuity into
adulthood (Bernstein et al., 1993; Gacono & Meloy, 1994; Lewinsohn et al., 1997; Westen
et al., 1990b; Westen & Chang, 2000). Despite this evidence, until recently, little research

appears to have been conducted on the underlying working representations of clinical or
non-clinical adolescent samples (Westen & Chang, 2000).
11
2.2. DILEMMAS IN DIAGNOSIS OF JUVENILE PSYCHOPATHOLOGY
The APA's diagnostic criteria for Conduct Disorder and Oppositional Defiant Disorder
almost exclusively focus on a behavioural checklist to the exclusion of personality
variables, hinting at the poverty of material on this subject (American Psychiatric
Association, 2000; Hare, Hart & Harpur, 1999). For instance, Conduct Disorder criteria
include aggression to people and animals, destruction of property, deceitfulness or theft,
and a serious violation of the rules (American Psychiatric Association, 2000). Widespread
criticism has been levelled at this means of categorisation, including a lack of empirical
proof of categories and criteria, excessively high rates of co-morbidity suggesting a lack of
discriminant validity, artificial dichotomization into present/absent, the inability to weigh
criteria, failure to consider personality strengths, the omission of a range of personality
pathology found in patients, an arbitrary categorical conceptualization in place of a
continuum approach, limited clinical utility and poor validity. (Gacono & Meloy, 1994;
Hare, Hart & Harpur, 1999; Westen & Shedler 1999; Westen & Chang, 2000). In addition,
an externally based and categorical means of evaluation, while arguably practical, ignores
the highly heterogeneous nature of adolescent Conduct Disorder (Gacono & Meloy, 1994;
Westen & Chang, 2000). For instance, it has been argued that adolescents diagnosed with
Conduct Disorder may demonstrate personality characteristics ranging from callous,
remorseless and psychopathic, to high-functioning, and able to maintain loving and intimate
relationships. This essential distinction has been identified as distinguishing serious

character problems from underlying personality conflict (Gacono & Meloy, 1994; Westen
& Chang, 2000).
12
Sugarman (1980) seems to have identified this anomaly in Rorschach protocols of severe
Conduct Disordered individuals many years previously, distinguishing between adolescent
borderline pathology and adolescent borderline features, with the latter grouping
demonstrating less severely disturbed responses. The identification and vital distinction
between adolescent-limited and life-course persistent antisocial behaviour further
emphasises the importance of unravelling essential differences within the broad category of
Conduct Disorder. In essence, current diagnostic criteria fail to distinguish between
psychopathic adolescents and those whose actions mask less disturbed personality features
such as depression, neurotic conflicts, and other normative and non-normative adolescent
concerns, thereby severely limiting treatment and recommendations (Moffitt, 1993; Westen
& Chang, 2000).
The inherent difficulty in diagnosing adolescent personality pathology according to APA
criteria has resulted in juvenile “character problems” being relegated to Axis 1, suggestive
of a hopeful developmental flux (Gacono & Meloy, 1994: 46). At 18 years of age,
Personality Disorder diagnoses appear on Axis 2, reflective of their chronicity and the
stability of character traits. In addition, the predictable movement of 25% of Conduct
Disorder diagnoses in childhood to Antisocial Personality Disorder at 18 years of age does
little to unravel underlying psychological attributes and offers the clinician negligible
therapeutic utility. And while only approximately 8% of Conduct Disordered children are
estimated to be at risk for primary psychopathy later in life, evidence has shown the

13
extreme and unremitting frequency of crime committed by a small number of psychopathic
males (Gacono & Meloy, 1994; Moffitt, 1993; Sadock & Sadock, 2003). Raine et al.
(1990), have gone so far as suggesting a subgroup of adolescents with clinical and
neuropsychological features similar to those found in adult psychopaths. The possibility of
identifying the intrapsychic risk factors of adolescents vulnerable to developing
psychopathy thus seems worthy of attention.
2.3. RESEARCH INTO ADOLESCENT AND ADULT PSYCHOPATHOLOGY
2.3.1. Fantasy & Aggression
The primary theme of projective research into violent juveniles in the 1950's and 1960's
seems to have centred around rival theories in explaining the role of fantasy in aggressive
behaviour (Keltikangas-Jarvinen, 1982). Direct Expression Theory proposed that
individuals with hostile fantasies were more likely to engage in aggressive and damaging
behaviour. Accordingly, individuals with aggressive fantasies were understood to enact
their internal worlds through violent behaviour (Mussen & Naylor, 1954; Kagan; 1956). In
contrast, Substitute Expression Theory suggested aggressive fantasies to have a substituting
or defensive function, rendering hostile impulses manageable through the role of symbolic
satisfaction (Murray, 1943; Sanford et al., 1943; Feshback, 1955; Keltikangas-Jarvinen,
1982). In addition, anxiety was suggested to play an inhibiting role in the overt expression
of aggressive thoughts (Kagan; 1956). Cowden et al. (1969) introduced a slightly different
perspective, suggesting assaultive delinquents lack a symbolic function and consequently
act out their hostile feelings. Much of the research during this period thus focused on

14
proving or disproving the relationship between fantasized and overt aggression.
2.3.2. Fantasy, Aggression & Punishment
An evolvement of the above theorisations concerned the relationship between aggressive
behaviour, fantasies of aggression and imaginings of external or internal punishment
(Kagan, 1956; Purcell, 1956; Lesser, 1957). Purcell (1956), using TAT responses, found
elevated themes of aggression among antisocial draftees and lowered themes of internal and
external punishment. In contrast, Shore et al. (1964) demonstrated increased internal
punishment or guilt in TAT protocols of delinquent adolescents. Lesser (1958), in a similar
vein, found themes of aggression to be associated with narratives of external punishment.
The differences in these findings seem in part attributable to the varying ages of subjects
and loose definitions of antisocial behaviour (Schaefer & Norman, 1967). Schaeffer &
Norman (1967: 237), in a comparative study of “anti-social” male juveniles and normals
using the TAT and other projective tests, found similar themes of aggression among both
groups, although the “anti-social” group followed violent narratives with depictions of
external punishment. It was thus speculated that punishment by external authorities would
reinforce defensiveness and reduce guilt related to anti-social behaviour.
2.3.3. Rorschach Studies
A number of studies assessing personality structure on the Rorschach have contributed
towards a greater understanding of the internal dynamics underpinning the personality
structure of violent males (Sugarman, 1980; Gacono & Meloy, 1992; Gacono & Meloy,

15
1993; Gacono & Meloy, 1994; Gacono et al., 2000). Sugarman (1980) identified a
preoccupation with maintaining self-other boundaries, part-object attachments,
idiosyncratic qualities, primitive modes of relating, and aggressive imagery among severely
Conduct Disordered adolescents using Rorschach protocols. Gacono and Meloy (1994)
administered Rorschach protocols to 100 Conduct Disordered adolescents. Results revealed
no expectation of co-operative interactions with others, shifting identifications between
aggressor and victim, social isolation, lowered and fantasy-based relational attitudes, partobject
representations, a primitive relating response, narcissistic mirroring, boundary
disturbance, violent symbiosis, poor maternal object identification, and absence of a
capacity for attachment or affect hunger. Overall, a chronic emotional detachment was
noted.
A case study of an adult with a diagnosis of Conduct Disorder, Borderline Personality
Disorder with Narcissistic features and a history of robbery and homicide revealed a strong
identification with the aggressor, secondary identification with the victim, a primary object
relation marked by violent attachment, poor modulation of affect and absence of unpleasant
emotion, fusion of aggressive and sexual drives in the context of the self as an injured
object, borderline personality organization and psychopathic character formation (Gacono
& Meloy, 1993). Psychopathic character formation was understood to entail chronic anger,
emotional constraint while incarcerated, identification with the aggressor, a sense of
omnipotence, perceptual unconventionality, relatedness to others as part-objects,
unmodulated affect and severely impaired reality testing (Gacono & Meloy, 1992).
Aggression was understood to be managed by the defensive operations of devaluation,
projection and projective identification, while splitting and denial facilitated the use of

16
sadism to achieve the conversion of passive into active, the law of revenge, momentary
pleasure through the infliction of pain, the omnipotent control of the aggressor, and the
projection of the injured self (Gacono & Meloy, 1993).
Gacono et al. (2000) innovatively attempted to ascertain the personality attributes of
psychopaths, sexual homicide perpetrators and paedophiles. Results indicated reality testing
deficits, pathological narcissism, and formal thought disorder among all three groupings.
The psychopathic group demonstrated little interest in others and a lack of ego-dystonic
internal conflict. Both the sexual homicide perpetrators and paedophiles showed an interest
in others and an exaggeratedly dysphoric internal world. In addition, sexual homicide
perpetrators were characterised by obsessional thinking and a denial of their wish for
contact with others as a consequence of their pathological narcissism.
2.3.4. Conduct Disorder, Psychopathy & Personality Disorder
Valuable research has explored the relationship between psychopathy, as measured by the
Revised Psychopathy Checklist (PCL-R), delinquent behaviours, Conduct Disorder, and
Personality Disorders (McManus et al.,1984; Hare et al., 1990; Myers et al., 1995). The
PCL-R is a 20-item instrument covering the following areas: superficial charm, grandiose
sense of self-worth, need for stimulation, pathological lying, conning/manipulativeness,
lack of remorse or guilt, shallow affect, callousness/lack of empathy, parasitic lifestyle,
poor behavioural controls, promiscuous sexual behaviour, early behavioural problems, lack
of realistic long-term goals, impulsivity, irresponsibility, failure to accept responsibility for
actions, many short-term marital relationships, juvenile delinquency, revocation of

17
conditional release, and criminal versatility (Myers et al., 1995). Previous studies among
adults had found positive correlations between PCL-R scores and Antisocial, Narcissistic,
and Histrionic Personality Disorders (Hare et al., 1990). Myers et al. (1995), using the
Structured Interview for Personality Disorders-Revised (SIDP-R) and the PCL-R in a study
of thirty adolescent inpatients, found significant relationships between elevated PCL-R
psychopathy scores and delinquent behaviours, Conduct Disorder and Narcissistic
Personality Disorder. This may however be in part due to the overlay between these
disorders with respect to feelings of entitlement, a need for attention, lack of empathy,
exploitativeness, and a grandiose sense of self.
Using the PCL-R, Gacono & Meloy (1990) compared a group of severe (PCL-R>30) and
moderate (PCL-R30) and nonpsychopathic (PCL-R

18
Conduct Disorder, Borderline and Narcissistic Personality Disorders. These findings
indicate the heterogeneous nature of Conduct Disorder and Psychopathy, as well as the
importance of assessing personality characteristics due to the very different origins and
manifestations of these diagnoses.
2.3.5. Hostile Attributional Bias
Social cognition theorists and researchers have made invaluable contributions in the
unravelling of aggressive children's personality characteristics. Much of the research has
centred around demonstrating the association between a child's interpretation of a peer's
intention, referred to as “attribution behaviour” and the child's response (Dodge &
Somberg, 1987: 214). For instance, a child who perceived a peer to have acted with hostile
intent would feel justified in retaliating aggressively, whereas a child who believed the peer
to have acted with benign motivation would react without hostility (Dodge et al., 1984;
Rule et al., 1974). Aggressive children were shown to interpret ambiguous conflict
situations with increased negative attributions, thereby resulting in aggressive responses.
(Nasby et al., 1979; Dodge et al., 1984).
In a study involving the introduction of a potential threat of conflict among 32 aggressive
and 33 non-aggressive boys between the ages of 8 to 10 years, Dodge & Somberg (1987)
discovered the aggressive boys to have less skill in accurately interpreting peer intentions,
lowered predictability of response and increased hostile attributions. These hostile
attributions were understood to be largely unconscious “cognitive structures”, or templates,
existing as “organised representations of prior experience” that processed information via

19
screening and coding (Dodge & Somberg, 1987: 221). The stimulation of these cognitive
structures was understood to evoke negative affective states, prompting the defensive
mechanism of externalised blame so as to to relieve internal discomfort and thereby justify
retaliatory action. The stage was thus set for further exploration of the cognitive and
affective structures of violent juveniles.
2.3.6. The Thematic Apperception Test (TAT) & The Social Cognitions & Object
Relations Scale (SCORS)
Much of the recent research on the personality characteristics of clinical adolescent samples
has involved the use of a multidimensional measure developed by Westen (1985), the
SCORS, for the clinical assessment of object relations using the TAT. The measure is
derived from clinical experience, object relations theory and research, extensive
investigations in developmental social cognition, and previous efforts to measure related
dimensions with the TAT and other scales (Westen, 1991b; Westen et al., 1985; Westen et
al., 1990b). The scale has been extensively validated in both clinical and nonclinical
populations (Ornduff et al., 1994; Westen, 1991b; Westen et al., 1991a). A brief outline of
the scale will follow so as to facilitate discussion of recent research findings (please refer to
Appendix A for Synopsis of Scales).
2.3.6.1. Complexity of Representations
In an innovative synthesis of various object relations theorists, Westen (1985) developed
four interdependent cognitive and affective dimensions of internal working representations,

20
each with five levels ranging from least to most mature levels of functioning. The
complexity of representations dimension measures the extent to which others are viewed as
multidimensional beings with complex motives and subjective experiences. Levels of
functioning range from a poor differentiation of the multidimensionality of self and others –
a profound egocentric embeddedness in one's own point of view – to a rich, complex, subtle
and psychologically-minded understanding of personality and subjective experience.
Westen (1989: 338) illustrates this dimension in a Borderline patient's inability to
distinguish an aspect of her “problematic” functioning from her entire self; an essential
conflating of parts with the whole. An extremely egocentric and poorly bounded sense of
identity, hypothesised as pathological, was also found among sexually abused female
adolescents. This was presumed to relate to the cognitive disruption and identity
disturbance produced by recurring experiences of overwhelming affect (Westen et al.,
1990e).
2.3.6.2. Affect Tone
The affect tone of relationship schemas dimension refers to expectations and attributions
concerning the affective quality of people, relationships, and the world. This dimension
ranges from a malevolent, threatening, painful and capricious viewpoint with little hope of
comfort or kindness between people, to an expectation of relationships as benign, enriching,
positive and consistent. For example, Ornduff et al. (1994) found a negative and punitive
affect tone in the interpersonal expectations of sexually abused children, while Westen et
al. (1990e) found a malevolent object world among sexually abused female adolescents.
Interestingly, the sexual abuse primarily occurred in latency, suggesting the influence of

21
post-oedipal events on object relations (Westen et al., 1990e).
2.3.6.3. Capacity for Emotional Investment in Relationships
The capacity for emotional investment in relationships and moral standards dimension
measures the degree to which others are treated as ends rather than means. Lower levels of
functioning are characterised by a need-gratifying and self-soothing orientation towards the
social world, while more mature scores indicate a commitment to relationships based on
values such as mutual sharing, interdependence and respect for others. Research indicated
Borderline patients to characteristically demonstrate lowered levels of functioning on this
dimension, often investing “too much too soon” (Westen, 1989: 340).
2.3.6.4. Understanding of Social Causality
The understanding of social causality dimension measures the comprehension of causes
underlying the actions, thoughts, and feelings of others. Levels range from idiosyncratic,
illogical and distorted, to complex, accurate and abstract with an understanding of possible
internal motivations. Severely personality disordered patients have shown low levels of
functioning on this dimension, indicating illogical and innacurate attributions regarding the
causes of behaviour (Westen, 1991c).

22
2.3.6.5. Research with the TAT & SCORS
2.3.6.5.1. Adolescent Object Relations
In a groundbreaking piece of research, Westen et al. (1991a) demonstrated different levels
of object relations between children in 9 th
(approximately 14-15 years) and 12 th
(approximately 17-18 years) grades. 12 th graders were found to demonstrate higher mean
scores for complexity of representations, emotional investment, and social causality than 9 th
graders (Westen et al., 1991a). Interestingly, the affect tone of relationship paradigms
showed a random pattern, refuting the notion of a normative progression from malevolent
to benevolent views of self and others. These findings demonstrate the importance of
considering adolescent object relations as multi-faceted, as opposed to unitary, and
developing beyond pre-oedipal years (Westen et al., 1991a).
2.3.6.5.2. Adolescent Borderline Object Relations
Research on object relational patterns of clinical adolescent populations is unfortunately
still in its infancy, reflected in the limited availability of data (Westen & Chang, 2000).
Using TAT material, Westen et al. (1990a) compared the object relations of adolescent
Borderline inpatients (diagnosed with the well-validated Diagnostic Interview for
Borderlines(DIB)) with a non-Borderline adolescent psychiatric comparison group (DSM
III diagnoses included Mood Disorders, Eating Disorders and non-Borderline Personality
Disorders) and normal adolescent subjects (screened using a questionnaire and the DIB).
Adolescent Borderline inpatients demonstrated significantly lower-level functioning than

23
psychiatric and normal comparison subjects on affect tone, indicating a greater malevolent
object world and view of relationships. They also evidenced the lowest mean score on
emotional investment, reflecting a need-gratifying orientation in relationships. Both patient
samples scored significantly lower than normal comparison subjects on social causality,
indicating greater idiosyncratic and illogical attributions of causality in understanding
human interactions. Borderline subjects also evidenced more pathological responses, scored
as Level 1 responses, on all scales, reflective of a poor differentiation of representations, a
malevolent object world, a need-gratifying orientation towards others, and grossly illogical
attributions. Most surprisingly, Borderline subjects evidenced the highest mean score on
complexity of representations. These results suggest adolescent Borderline representations
of others to be relatively normal, although they may at times be shallow and poorly
bounded. It was speculated that the nature of Borderline pathology could at times incline
such adolescents to over-attribute complex motives and enduring traits to others, while
simultaneously struggling with object differentiation in moments of affective arousal
(Westen et al., 1990a).
2.3.6.5.3. Adult Borderline Object Relations
In further research, Westen et al. (1990c) compared the object relational patterns of adult
Borderline patients with major depressives and normal subjects using the TAT and SCORS.
Results indicated borderlines to have lower mean scores on affect tone and emotional
investment, and to produce more pathological responses on all scales in comparison to the
major depressive group. The Borderline group also produced lower mean scores and greater
pathological responses on every scale when compared to normals. Westen et al. (1990b)

24
have further investigated the developmental differences between Borderline adolescents
(DIB diagnosis) and adults (DSM III diagnosis). Results indicated similar object relational
disturbances among both groups relative to normal and clinical comparison groups, even
though certain aspects of object-relational patterns appear to mature between adolescence
and adulthood (Westen & Chang, 2000).
2.3.6.5.4. Developmental History & Object Relations
In a most interesting research study exploring the relationship between actual
developmental history events, as assessed by chart review, and object relational patterns of
female adolescent inpatients (diagnosed with Borderline Personality Disorder, Mood
Disorder, Anorexia Nervosa, Bulimia, and other Personality Disorders) using the TAT and
SCORS, Westen et al. (1990e) identified the maternal relationship as crucial in the shaping
of object relational pathology. Maternal psychiatric illness was found to best predict
pathological object relations on all dimensions except complexity of representations, while
maternal separations, alcohol abuse, and mother surrogates appeared to negatively impact
on all object relational dimensions. In addition, early separations were associated with
greater pathological levels of object relations. Surprisingly, maternal physical abuse and
inappropriate parental behaviour were both associated with more high-level responses on
complexity of representations. A myriad maternal difficulties, pre-oedipal risk factors,
disrupted attachments and sexual abuse were associated with lowered affect tone, indicating
difficulty with the formation and retrieval of benign and benevolent representations. The
duration of sexual abuse correlated with Level 1 scores on each of the dimensions, as well
as significantly suppressed scores on complexity of representations. A history of neglect

25
was associated with illogical attributions and a lowered affect tone. While no single preoedipal
risk factor distinguished Borderline from non-Borderline psychiatric patients, the
amount of risk factors predicted diagnosis, was highly correlated with affect tone and
capacity for emotional investment, and slightly less significantly correlated with complexity
of representations and understanding of social causality. Most significantly, a near zero
correlation existed between post-oedipal risk factors and the various dimensions of object
relations, highlighting the importance of the pre-oedipal period in childhood development.
The significance of the above study lies in its ability to demonstrate the interface between
actual reality and psychic reality, with the type of early trauma significantly impacting on
later object relational patterns. It still remains to be ascertained whether these internal
patterns and expectations of others remain static or are momentarily activated under
conditions of strong negative affective arousal, thereby disrupting cognitive processes and
evoking archaic and malevolent representations.
2.3.6.5.5. Object Relations of a Serial Sexual Homicide Offender
Porcerelli et al. (2001), in one of the few studies investigating the object relations of an
incarcerated serial sexual homicide offender using the TAT and SCORS, uncovered a
malevolent affect tone and a need gratifying orientation towards others. Most interestingly,
scores on complexity of representations indicated an ability to view others in complex
ways, although a difficulty was demonstrated in maintaining consistent boundaries. This
was proposed to facilitate a projection of hated and rageful aspects of the self onto others,
thereby requiring the sadistic control and elimination of these malevolent external objects.

26
Westen (1991c: 441) attributes Antisocial Personality Disorders' elevated scores on
complexity of representations to their associated “psychological acuity” and interpersonal
cunningness, combined with a lack of affective-motivational investment in others. In a
similar vein, social causality was often understood in both complex and illogical ways,
suggestive of a disturbed and idiosyncratic understanding of cause and effect. Westen
(1991c) speculates on this distortion as reflecting the Antisocial Personality's awareness of
causal impact, while simultaneously enjoying the infliction of pain.
2.4. CONCLUSION
The link between object relational patterns and violent behaviour alluded to by Porcerelli et
al. (2001) is tenuous, with a theoretical link more established than an empirical one.
Gacono & Meloy (1990), in their Rorschach studies of violent offenders, do argue for
similar characterological features among violent offenders. Understanding violent
behaviour requires a multidimensional understanding, including but not restricted to social,
political, economic, and psychological factors. Criminologists are however finding
inadequate explanation for the elevated levels of crime in South Africa, and particularly the
Western Cape, and even fewer explanations for increases in violent juvenile crime (Legett,
2004a). And while none of these factors can singularly predict violent behaviour,
understanding the object relational patterns of violent male juvenile offenders presents
itself as a worthwhile area of research into the intrapsychic factors that may render certain
juveniles susceptible to violent behaviour.

27
CHAPTER 3: METHODOLOGY
3.1. STUDY DESIGN
Much of the research on juvenile offenders and criminal activity occurs within a
behavioural perspective (Sadock & Sadock, 2003). Gacono & Meloy (1994) have
highlighted the importance of understanding personality factors underpinning criminal
behaviour. Object relations theory and methods of evaluation have successfully been
deployed in attempts to ascertain personality characteristics and prototypal patterns of
violent offenders (Gacono & Meloy, 1994; Porcerelli et al., 2001). The present study
adopted an object relations theoretical framework as its conceptual lens. The research
design was composed of a number of instruments, including the Thematic Apperception
Test (TAT) and the Social Cognitions & Object Relations Scale (SCORS). The purpose of
the design was to assess and compare object relational patterns from TAT narratives.
3.2. PROJECTIVE TESTING
A burgeoning interest in the role of unconscious wishes, motivations and internal conflicts
prompted the development of projective assessment methods in the early 1900's. The first
projective test dates back to Herman Rorschach's experimentation with ambiguous inkblots
in approximately 1910 (Atkinson et al., 1985). The Thematic Apperception Test (TAT)
represented a further significant contribution to the field of projective testing. The TAT,
consisting of thirty ambiguous pictures and one blank card, was developed by Murray and
Morgan at the Harvard Psychological Clinic in 1943 (Sugarman, 2000). The fundamental

28
tenet of projective tests rests on the presentation of ambiguous stimuli, be they in the form
of Rorschach inkblots or pictures from the TAT. The ambiguous nature of the stimulus is
understood to invite the participant to organise and associate the material, thereby
presumably reflecting implicit motives (Masling, 1997).
Interpretation of psychometric tests initially rested on the thematic analysis of content. This
means of analysis was to acquire a dubious reputation due to the plethora of cookbook
interpretations offering "wildly specific and detailed dynamic interpretations", as well as the
recognition of the importance of structural aspects of the personality (Sugarman, 2000: 6).
An emphasis on validity and reliability ushered the development of quantitative means of
projective analysis. Actuarial methods of assessing projective results have dominated
published literature for the last thirty years or so, with a number of scoring schemes
appearing for the Rorchach (Masling, 1997). Test scores are understood to give precision to
underlying theoretical components and allow a comparison with other measured personality
constructs, thereby facilitating an evaluation of the structural aspects of the personality
(Appelbaum & Mayam, 1974). The analysis of content has not however been relegated to
the archives. Complementary approaches attempting to assess both structural dimensions
and thematic qualities of object relations have emerged (Sugarman, 2000). Westen's (1985)
SCORS represents an attempt to integrate these two perspectives, bringing both greater
reliability, validity and a heuristic value to projective testing.

29
3.3. METHOD
3.3.1. Characteristics of the Sample
Eight participants were selected for the research study using the following criteria: (1) they
were male, (2) under 18 years of age, (3) had been found guilty of committing a violent
crime by a South African court of law, (4) and were serving a sentence. The 8 male juvenile
offenders were incarcerated for violent crimes at De Novo Youth Care Centre in
Kraaifontein and Eureka Youth Care Centre in Rawsonville. Potential participants were
identified with the assistance of the Social Workers at both institutions and ranged from
fifteen to seventeen years of age. Participants had all been found guilty of violent crimes,
including murder, man-slaughter, rape, and armed robbery, thus constituting a selected
convenient sample.
3.3.2. Recruitment & Procedure
Written permission for the research study was obtained from the Western Cape Education
Department, followed by written and telephonic communications with the Principals of De
Novo and Eureka Youth Care Centres. Informed and voluntary consent of the participants
was obtained, as is recommended in socially sensitive research with juveniles (Jenkins-Hall
& Osborn, 1994). The purpose of the research and the procedure to be followed was
explained to all the potential participants by a staff member, followed by the researcher.
Individual confidentiality was guaranteed and participants were advised of their right to
refuse participation in the study. It was further made explicit to the potential participants

30
that participation or non-participation in the research study would in no way advantage or
disadvantage them. The research was emphasised to be independent of the management of
the institution and as having no impact on their periods of sentencing. Potential participants
were informed of their right to withdraw at any stage of the interview and encouraged to
avail themselves of the limited counselling available should the interviews upset an
emotional equilibrium or result in any emotional discontent.
Voluntary participation was requested by the researcher. None of the identified participants
declined to participate, although one interview at De Novo Youth Care Centre was
terminated by the researcher after the participant declared an inability to construct a story,
claiming it impossible to know what characters in the TAT cards might be thinking or
feeling. Each of the interviews was conducted one-to-one and face-to-face in a private
office at either De Novo or Eureka Youth Care Centres. All of the interviews were taperecorded
after receiving permission to do so by the participants. All participants, apart from
the one terminated interview, participated enthusiastically and appeared to enjoy the
opportunity to construct their own stories. Many of the participants, after having narrated a
story, emphasised a parallel between their constructed narratives and personal life
experiences. Counselling was recommended by the interviewer in a number of instances.
In addition, the interviewer reviewed all available documents, if possible prior to the
interview, to verify the nature of the offences and court verdict. These included legal
dockets, police reports, victim and witness statements and Social Worker reports.

31
3.3.3. Research Tool
The TAT is a psychological projective technique developed at the Harvard Psychological
Clinic in 1943 (Wesley, 2002). The original test comprised 30 picture cards and 1 blank
card that act as stimuli for evoking stories from participants (Osborne, 1996). Participants
are usually asked to tell a story about the picture, including what is happening in the picture,
what led up to it, the outcome, and what the characters are thinking and feeling (Westen et
al.,1990a). Brief prompts such as "How does this story end?", "Tell me the thoughts and
feelings of the characters in this story", or probes for unclear or confusing information may
be used (Osborne, 1996: 4; Westen, 1985). This method of administration is endorsed by
Bellak et al. (1949). The ambiguous nature of the interpersonal situations depicted on the
cards provides an excellent source for eliciting underlying cognitive and affective structures
and themes of meaning. These are in turn understood to reveal how relations and
interactions are experienced, constructed, and made sense of (Osborne, 1996; Ornduff et
al., 1994; Westen et al., 1990a; Westen et al., 1990b).
The TAT is a well established tool for use with adolescents - 63% of clinicians reported its
usage, making it the 4 th most popular psychological instrument used in standard test
batteries with this age group. This has been attributed to the test's creative quality, as well
as its indirect, informal, and non-structured approach (Osborne, 1996). While the bias of
verbal productivity was initially of concern, it has conclusively been shown that the number
of words produced does not change the pattern of findings or levels of significance (Westen
et al., 1990a).

32
The TAT method in this study follows the procedure outlined by Westen (1985).
Accordingly, ten TAT cards were chosen so as to yield internal consistency. The following
pre-selected TAT cards were administered in a standerdized sequence: Cards 1, 6BM, 12M,
7BM, 4, 3GF, 18BM, 8BM, 13MF, and 14. The above cards were chosen due to their
elicitation of relational and emotive themes (Bellak, 1949). Standard clinical administration
was followed and interviews lasted approximately 1 hour each.
3.3.4. Data Analysis
TAT narratives were professionally transcribed and thereafter verified by the researcher.
The protocols were scored using Westen's (1985) SCORS scale that has been extensively
validated in both clinical and nonclinical populations. This involved analysis of the four
previously described dimensions of object relatedness (Complexity of Representations,
Affect Tone, Capacity for Emotional Investment, and Understanding of Social Causality)
according to a five level scoring system ranging from least to most mature levels of
functioning, with each level including numerous scoring criteria. Assessment of these
dimensions required extensive and labour-intensive training with the coding manual,
including: (1) familiarisation with the aims and theoretical background of the measure, (2)
multiple readings of the detailed scoring criteria for each of the five levels, (3) scoring two
practice protocols and (4) scoring one reliability protocol. In total, this amounted to scoring
seventy five sample TAT stories against scoring rationales and corresponding scores
(Westen, 1985). All sample narratives were analysed on a single dimension at a time, as
advocated by Westen (1985), due to the complexity of the scales and so as to yield internal
consistency. Analysis of the four dimensions necessitated the scoring of these seventy five

33
sample narratives four times, elevating the number of stories scored to three hundred.
The eight participants produced lengthy responses to the TAT cards. A total of eighty
narratives, ranging from twenty to fifty typed A4 pages, were produced. Seven of the eight
interviews were conducted in Afrikaans. These were translated into English by the
researcher, a second language Afrikaans-speaker, prior to scoring so as to ensure a common
language base. Stories were analysed and scored on each of the four dimensions separately,
totalling three hundred and twenty narratives. The large number and length of narratives,
translation, and an intricate scoring manual requiring extensive familiarisation and training
contributed to an inordinate amount of time spent on data analysis. Although SCORS
details a quantitative score, it also provides insight into the qualitative dimensions of object
relation patterns. Results were therefore examined according to descriptive data and
illustrated with qualitative material.
3.4. LIMITATIONS & DELIMITATIONS
Delimitations included restricting the participants to eight juvenile males incarcerated for
violent crimes as the majority of violent crimes appear to be committed by males (DCS,
2002; DCS, 2003; DCS, 2004; DCS, 2005). The sample size was restricted to eight
participants due to the lengthy scoring procedure of SCORS, as well as wishing to remain
within the parameters of a Clinical Masters in Psychology mini-thesis. The limited number
of participants does however limit the generalizability of findings. No intergroup or crossdimensional
statistical comparisons were possible due to the sample size. As a result the
data was not subjected to a t-test or chi square analyses. More specific delimitations may

have included grouping participants according to socio-economic background, race, and/or
an intelligence score.
34
3.5. RESEARCH HYPOTHESIS
It was hypothesised that violent male juvenile offenders would demonstrate low scores on
affect tone and capacity for emotional investment, while displaying relative elevations on
complexity of representations and understanding of causality.

35
CHAPTER 4: RESULTS & DISCUSSION
4.1. DESCRIPTIVE STATISTICS
Table 1 depicts the object relations scores of the participants. The table shows the number
of subjects, minimum and maximum total scores, mean scores and the standard deviation.
As can be seen from the table, the initial hypothesis was supported: mean scores for
complexity of representations and social causality demonstrated relative elevations when
compared to affect tone and capacity for emotional investment. Participants thus
demonstrated elevated scores on the cognitive dimensions (complexity of representations
and understanding of causality) relative to the scores on the affective dimensions (affect
tone and capacity for emotional investment).
Descriptive Statistics
Complexity of
representations
Affective quality of
representations
Investment in
relationships
Social causality
Valid N (listwise)
N Minimum Maximum Mean Std. Deviation
8 18 45 32.38 8.911
8 18 33 22.63 4.984
8 16 20 18.00 1.604
8 20 37 26.38 6.255
8
Table 1

36
4.2. MEAN SCORES ACROSS ALL DIMENSIONS
40
30
20
Mean
10
Complexity
Affective tone
Emotional Investment
Social causality
Figure 1: Mean scores across all dimensions
Figure 1 depicts the mean scores across all dimensions. As can be seen from the table,
participants scored highest on complexity of representations, followed by social causality,
affect tone and emotional investment.
4.2.1. Complexity of Representations: Mean Scores
Scores for complexity averaged Level 3 (please refer to Figure 1), reflecting an ability to
“make inferences about subjective states in addition to focusing on behaviour” (Westen,
1985: 4). The understanding of psychological processes and subjective states at this level of
object relations is however rather superficial, unidimensional, generalised and lacking in

37
subtlety. People are therefore generally not understood or expected to act out of character or
experience psychological tensions. Level 3 scores nevertheless represent an average level of
maturation. The participants thus demonstrated an ability to infer basic psychological states
of others, suggestive of the social acuity characteristic of the psychopathic personality
structure (Gacono & Meloy, 1994; Westen & Chang, 2000). A number of recurring themes
occurred on this dimension.
4.2.1.1. TAT excerpt: Simple traits - Just Bad
Inferences about subjective states generally depicted simple and malevolent character traits,
as reflected in MW's response to Card 1:
"He is heartsore, because he broke the thing. He feels unhappy and heartsore and down.
Now he's sitting now and thinking what his mother's going to do with him; mother or
father, because he broke the thing".
Likewise, MW, on Card 6BM, depicted the following scenario following a mother's
discovery of her son's involvement in illegal activities:
"Now he's thinking, she doesn't want to talk you see, she looks cross. The mother is cross
with the boy and the boy, he doesn't worry about the mother now. He doesn't feel so happy.
He doesn't feel so happy. He feels unhappy, and his mother is totally unhappy about what
he did."

38
In a similar vein, RP, on Card 3GF, described unidimensional characteristics of a woman
feeling "heartsore", "bad" and "tearful" following the ejection of her unfaithful husband,
while RT described a woman as simply "heartsore" following her husband's infidelity on
the same card.
4.2.1.1.1 Commentary on TAT excerpt: Simple traits - Just Bad
While participants demonstrated insight into the psychological complexity of others, these
were generally in simple and malevolent ways. This echoes Westen's (1990d) finding of
Borderline patients demonstrating highest-level attributions on human figures depicted as
malevolent, while depressives and controls did so on benevolent figures. The ability to infer
psychological states, however superficial, is likely to support the non-egocentric capacity to
"read" and manipulate others, characteristic of the psychopathic personality structure. In
addition, in moments of duress or object provocation, the psychological complexity of
others may be simplified into a malevolent representation, prompting a violent outlash in an
attempt to quell internal discomfort and promote object cohesion.
4.2.1.2. TAT excerpt: Antithetical traits - The Good & the Bad
Participants demonstrated insight into the mental lives of the characters depicted, although
these were often characterised in simple and antithetical ways. For instance, in response to
Card 12M, KF described an overwhelmingly hostile relationship between an "evil dark
man" and a boy described as a "saviour":

39
"He's some evil dark person, who's got evil man ... evil powers. He's trying to possess him;
to bring the evil spirit inside of him ... so that he can use his powers for bad. The boy wakes
up. The man gets scared or something and the boy gets cross and maybe he finds his
unhidden power and uses it against him and the man dies or blows up, or turns into dust or
something. And the boy fulfils his destiny and saves the world or something from evil."
HB also depicted this dichotomised style of thinking in Card 8BM that involved a story of
murder, cannibilism and the trading of body parts:
"The boy's thinking yeah, I've got the bucks for killing this man, but there is guilt; he must
have guilt. He's thinking to himself yey this is wrong. Look what they doing to this man;
chopping him up. The men just want what they want. They're not worried about other
people. They don't have feelings. All the sin and all the bad things they have done makes
them so blind they don't feel anymore. They have a cold heart. They after money or power."
4.2.1.2.1. Commentary on TAT excerpt: Antithetical traits - The Good & the Bad
The failure to integrate ambivalent representations and the continuation of split
representations beyond oedipal years, perhaps in an attempt to protect the object from rage,
reflects Kernberg's (1975) borderline pathology organization and an egocentrism in the
understanding of others' thoughts. This style of split representation of self and other, as
previously shown in Rorschach studies (Gacono & Meloy, 1994), is likely to facilitate the
projection of hated parts of the self as the complex composition of others is reduced to an
all-or-nothing disposition. It may well be that during periods of intense emotional arousal,

40
the complex representations of others are overwhelmed by the ascendancy of malevolent
representations. The escalated intensity of these malevolent representations is likely to
prompt projective defences in order to alleviate the pressure of a charged object world,
ultimately requiring the destruction of the external object so as to restore internal cohesion.
4.2.2. Complexity of representations: Individual Mean Scores
Participants varied in their mean scores on complexity of representations, as can be seen in
Figure 2. Four of the participants scored Level 3 or below, two between Levels 3 and 4,
and two between Levels 4 and 5. In addition, participant AM demonstrated multiple Level 1
and 2 responses. Examples of high (Levels 4 and 5) and low (Levels 1 and 2) scores will be
given to illustrate the full range of responses.
Value Complexity of representations
50
40
30
20
10
RP
RT
HB
RB
AM
RM
KF
MW
Respondent
Figure 2: Complexity of representations

41
4.2.2.1. TAT excerpt: Complex representations
Both KF and HB demonstrated a number of complex representations in their narratives. For
instance, on Card 13MF, KF described the murder and rape of a young woman by her
classmate and the response of her innocent partner:
"He doesn't know how to control it, he doesn't know what to do. He just covers his face in
shame and thinks maybe he came a little bit too late, maybe he could have stopped it. But
now he feels ashamed. That's why he's covering his face. He's ashamed and he's angry.
He's busy thinking who could do such a thing. Why wasn't I here. It's all on him, but its not
his fault."
HB's complex and enduring understanding of dynamic psychological process is reflected in
the following narrative to Card 8BM that centres around an organ transplant and the
differing responses from team members:
"I think she gave the instructions. She actually made the decision. She said to them they
must take the heart out. She's thinking about her salary. She feels nothing. She just thinks
about the money. And I think she's thinking about the food on the table at home this
evening. She's going to sit back and put her feet on the table and drink a cup of tea. The
Doctor is thinking also just hey must I again cut someone's flesh. But the money is there. I
can't put food on my table without having money. He's thinking can't I get another job. But
he thinks, this is the only work he's good at. I don't think the other Doctor worries about
what he does cause he sees it like the person is already dead. We can cut and dig in him as

42
we want to. He really doesn't care. The Doctor gets her money, the one that gave the
orders. Sits at home with her feet on the coffee table. Sits and drinks tea. And I think the
assistant is in a bar. He's drunk. And this Doctor, is at home, this doctor, the older man,
helps his grandchildren with schoolwork, homework."
4.2.2.1.1. Commentary on TAT excerpt: Complex representations
HB's narrative reflects an ability to differentiate characters, ascribe varying motivations,
describe differing mental states and extend these traits in concrete and enduring ways. His
ability to view people in complex ways highlights some of the variation among violent male
juvenile offenders' personality structure. Thus while mean scores for complexity of
representations averaged at Level 3, both HB and KF stood out as scoring between levels 4
and 5. Interestingly, both these participants also showed elevated scores on the
understanding of causality (please refer to Figure 5), possibly reflecting their heightened
understanding of these intellectual dimensions. Nevertheless, HB was found guilty of
murder while KF was incarcerated for armed robbery. Thus despite an ability to infer the
psychological states of others in highly complex ways, both these participants failed to
consistently hold these representations, effectively reducing their victims into objects
requiring subjugation or eradication.
4.2.2.2. TAT excerpt: Pathological representations
In stark contrast to the above, AM often struggled to differentiate characters on Card 7BM,
depicting them as sharing the same thoughts and emotions:

"They were both got up to bad things. They were both gruesome. He's thinking the same.
They feel happy now because they now have a new life. They turned over a new leaf."
43
4.2.2.2.1. Commentary on TAT excerpt: Pathological representations
AM, in what Westen (1985) calls "twinning", struggled with defining characters in clearly
bounded ways. Persons were often described as sharing the same situations, thoughts,
feelings or intentions, reflecting a failure to recognise the independence and separateness of
characters. The inability to recognise the separateness of others is likely to facilitate
overidentification in moments of high emotional arousal or duress, prompting an impetus to
eradicate the perceived extensions of the self as a means of restoring and preserving less
threatening object relations.
These findings iterate the importance of understanding the heterogeneity of object relational
patterns within a grouping of violent male juvenile offenders (Gacono & Meloy, 1994;
Westen & Chang, 2000). Some of these individuals may therefore understand others in
multidimensional ways, while a number of them would appear to struggle with a confusion
of points of view. Thus while the present sample were generally able to make inferences
about psychological states, these did fluctuate from simple to highly complex.
4.2.3. Affect Tone: Mean Scores
Scores for affect tone averaged Level 2, indicating a predominantly hostile, empty and
capricious view of the world (please refer to Figure 1). Level 2 scores indicate tremendous

44
feelings of loneliness, coupled with experiences of others as unpleasant and uncaring
(Westen, 1985). The participants thus demonstrated a malevolent view of the world with
little expectation of interpersonal co-operation. A number of recurring themes occurred on
this dimension.
4.2.3.1. TAT excerpt: A Violent & Aggressive World
Participants often depicted hostile and aggressive characters existing in a threatening world
filled with victims and victimisers. For instance, KF, on card 13MF, while demonstrating a
psychological understanding of the antecedents related to violence, nevertheless depicted a
grossly malevolent world following the rape and murder of an innocent young man's
girlfriend:
" ... he has this guilt inside of him and then he becomes like a person that's always looking
for trouble with people, always trying to get all that anger out of him. And he takes it out
on other people. Like maybe he walks past someone, someone bumps into him. He got
maybe a knife in his pocket. Starts stabbing."
RP, also in response to card 13MF, depicted a vindictive view of the world:
"... and then he realised she was playing with his feelings. He was just going to kill her.
The children must now grow up without a mother and live on the street. They beg and sleep
in a bin."

AM described an unambiguously malevolent representation on card 8BM involving a
surgical operation on a soldier:
45
"Now the Sister does not want to help him. She's thinking, no wait, leave him, let him die.
She does not like him. He probably did something ugly or something hurtful that's now
sitting in her heart. It's probably an ugly word that he said to her. When he's better and
sees the Sister then he's going to say to her: 'yes, you, you pig', or something like that. 'You
did not want to help me'. Then the sister said: 'yes, I won't help you. You should have just
died, you pig'."
4.2.3.1.1. Commentary on TAT excerpt: A Violent & Aggressive World
The above seems to support Direct Expression theorists' position of hostile fantasies
reflecting a greater propensity to engage in violent and destructive behaviours through the
enactment of an aggressive internal world (Mussen & Naylor, 1954; Kagan; 1956). The
lowered and violent nature of affect tone similarly echoes Sugarman's (1980) discovery of
aggressive imagery among severely Conduct Disordered adolescents using Rorschach
protocols. The participants' strong and split identifications with both the aggressor and the
victim are common in primitive object relation patterns (Gacono & Meloy, 1994).
Identifications with the aggressor may indicate a desire for omnipotent control, while
secondary identifications with the victim possibly reveal an injured sense of self resulting in
a preoccupation with physical vulnerability.
It therefore seems feasible to propose that violent male juvenile offenders' aggressive

46
actions may in part be attributed to the immobilisation of the cognitive ability to infer the
psychological states of others by malevolent affect-based perceptions in periods of intense
emotional arousal. A desire to restore a semblance of omnipotent control may in all
likelihood prompt violent actions, thereby restoring a primary identification with the
aggressor and control over a malevolent world. The participants' relational world thus
appears to be construed less in terms of affectional ties and more in terms of skewed power
relations.
4.2.3.2. TAT excerpt: A Capricious & Painful World
Participants often depicted a capricious world involving the infliction of sudden and
unexpected hurtful actions. For instance, in response to Card 3GF, RT described a woman
as extremely "heartsore" following the unexpected termination of her marriage:
"He maybe went out to another woman and came home smelling of perfume. They fought.
Her husband left her. She's wondering: 'How can he do that. And he never did that before;
we got on so well together. He probably got bored of me and walked.' She feels very
heartsore at that moment. Standing in the door. Standing and crying 'cause her heart is
broken."
In a similar vein, HB described an attempted suicide following a man's discovery of his
wife's infidelity on Card 18BM:
"He opened the door and found his wife and another man in bed. And they're married

47
maybe five years or so and he can't take it. What he saw there goes to his head and from
his head to his heart and he tries to kill himself. He feels heartsore. He feels his heart is
broken. He asks himself the question: Why does his wife do it? He asks himself if the love
that he gave to her was not enough."
4.2.3.2.1. Commentary on TAT excerpt: A Capricious & Painful World
In line with research on Borderline patients (Westen 1990b), sexually abused female
adolescents (Ornduff et al., 1994), sexual homicide perpetrators and paedophiles (Gacono
et al., 2000), the participants were found to demonstrate similarly dysphoric internal
worlds. Violent male juvenile offenders do therefore not appear to demonstrate the chronic
emotional detachment characteristic of psychopathic character formation, but an object
world filled with painful, aggressive, capricious and threatening emotions (Gacono &
Meloy, 1994). It may further be speculated that difficulty in regulating these affective
processes during times of duress may result in the treatment of others in ways likely to
evoke intense anger or desertion, thereby confirming their fantasies of a cruel world and
justifying retributive action.
4.2.4. Affect Tone: Individual Mean Scores
The majority of the participants scored below Level 3 on affect tone, as can be seen in
Figure 3. Three of the participants scored between Levels 1 and 2, four between Levels 2
and 3, and one between Levels 3 and 4. Excerpts from RM's responses will be shown to
illustrate higher level (Levels 4 and 5) responses on this dimension.

48
40
Value Affective quality of representations
30
20
10
RP
RT
HB
RB
AM
RM
KF
MW
Respondent
Figure 3: Mean Affect Tone
4.2.4.1. TAT excerpt: Variations in Affect Tone
RM, in response to card 7BM, described predominantly positive interactions between a
grandfather and his grandson:
"Looks to me like the man fought at work and came to tell his grandpa. And the grandpa
takes him by the neck and says to him: 'my son, don't worry. Everything will come right.
Just be strong'. And the grandpa sees how badly he (the grandson) feels cause he also
experienced something like this and it's very sore in your deepest insides."
Similarly, on Card 8BM, despite a schoolboy's witnessing of a murder, RM described a

49
number of supportive relationships:
"And his mother also saw that when he came out of school he didn't look like he always
looks. He looks otherwise. And his mother saw that evening when they were sitting and
eating at the dinner table that he's not speaking; he's sitting dead still. And his father asks
his mother what's going on with the child tonight ... And the next day there was someone
who came to speak at their school. And this man told them that if you've got something to
say to someone and you don't just trust anyone, you can go to your school teacher, or your
mother, or father or the police or someone you can trust. The meeting was finished and he
asked his teacher if he could talk to her one side."
4.2.4.1.1. Commentary on TAT excerpt: Variations in Affect Tone
While RM demonstrated an elevated score on affect tone in relation to other participants,
his responses were nevertheless characterised by affectively mixed representations of a
mildly negative tone (Westen, 1985). Thus even the variations in affect tone of relationship
paradigms indicated a dysphoric object world, lending further weight to the hypothesis that
states of negative emotional arousal may permit the ascendancy of a dysphoric, capricious
and painful affect tone that overwhelms cognitive dimensions and prompts vengeful
retaliation.
4.2.5. Capacity for Emotional Investment: Mean Scores
Scores for emotional investment averaged Level 2, indicating a pathologically narcissistic

50
orientation to the world (please refer to Figure 1). These low scores echo similar findings in
studies of Conduct Disordered adolescents, Psychopaths and Borderlines (Gacono &
Meloy, 1994; Gacono et al., 2000; Westen et al., 1990a). At this level, a basic awareness of
the conflict between the needs of the self and others is recognised, although the satisfaction
of one's own wishes remains paramount. An elementary distinction between right and
wrong is understood, although adherence to morals occurs primarily as a means to avoid
punishment. Friendships are viewed as interchangeable and usually occur within the context
of a shared activity (Westen, 1985). Participants described a number of recurring themes
on this dimension.
4.2.5.1. TAT excerpt: A Need Gratifying Orientation
Participants often described a need gratifying orientation towards others in their narratives,
as evidenced in RB's response to Card 13MF:
"He killed the woman. He's standing there and crying; looking on. 'My wife is now dead.
Where am I going to get another wife?'."
RB, in response to Card 3GF, described a narcissistic course of action following the beating
of a woman by her husband on suspicion of infidelity:
"The man leaves and feels nothing. He has nothing to do with the woman. He leaves the
woman just like that and goes looking for another wife to raise his children."

51
4.2.5.1.1. Commentary on TAT excerpt: A Need Gratifying Orientation
The above narratives demonstrate a pathological narcissism that has long been associated
with psychopathic character formation (Gacono & Meloy, 1994). Despite elevated scores on
the cognitive dimensions, low scores on the affective dimensions indicate an impoverished
emotional quality among the participants. In a circulatory fashion, a narcissistic orientation
would seem to foster deficits in the capacity for mutual relatedness, thereby limiting
meaningful interactions with others and encouraging the view of the world as unsafe and
unloving (Porcerelli et al., 2001).
It may be speculated that rageful and hated affective components become mobilised under
periods of duress or hyper-arousal, immobilising the cognitive ability to infer the
psychological states of others. Pathological levels of narcissism are likely to defeat any
residues of empathy in order to fulfil their need gratifying orientation, thereby assisting in
the fulfilment of grossly aggressive actions. The sample of violent male juvenile offenders
thus appears unable to cathect in others, thereby limiting intrusions into their fragile inner
world and simultaneously facilitating the sadistic control of others.
4.2.5.2. TAT excerpt: Violence with No Remorse
Participants' responses often included deleterious acts with no sense of remorse. HB, in
response to Card 4, described a character's rage at the approach of his girlfriend by another
man:

52
"He's thinking just about one thing. He's thinking about the person that did him wrong.
He's thinking he must get that person back. He can't help that he's cross. He feels cross. He
feels hateful. He just wants to kill."
RP also described a scene of impulsive aggression with no sense of remorse in response to
Card 18BM:
"He maybe stood now by the shop or came out of the shop and they watched him maybe
taking lots of money out of his pocket or buying. And when he comes out of the shop going
to his car then they hit him from behind. The man can die unless other people come quickly
to help him. Maybe he had a small child with him. Then they just take his child okay.
Kidnapping. Make it dead."
4.2.5.2.1. Commentary on TAT excerpt: Violence with No Remorse
Participants thus appeared to construe relationships primarily as vehicles for gratification
and security, with good and bad simply viewed in terms of the hedonistic implications of
action. The internalisation of moral standards, let alone abstract forms of moral reasoning,
are starkly absent. Others do not appear to be valued as independent entities, but simply as a
means of achieving personal satisfaction. It may well be that a recognition of the
independence and value of others represents an imminent threat to their fragile and
narcissistic sense of selves, prompting the sole investment in their own satisfaction. The
absence of moral reasoning, coupled with the drive for narcissistic self satisfaction and a
malevolently skewed perspective of the world are likely to congeal into a formidably

destructive force capable of overwhelming cognitively driven dimensions in moments of
distress, thereby prompting the perpetration of gruesome acts.
53
4.2.6. Capacity for Emotional Investment: Individual Mean Scores
All of the participants scored Level 2 or below on emotional investment, as can be seen in
Figure 4. No significant variations occurred across this dimension and it can be concluded
that all participants displayed an egocentric embeddedness (Westen, 1990d).
Value Investment in relationships
21
20
19
18
17
16
15
RP
RT
HB
RB
AM
RM
KF
MW
Respondent
Figure 4: Emotional Investment

54
4.2.7. Understanding of Social Causality: Mean Scores
Scores for social causality averaged Level 3, indicating a relatively simple psychological
understanding of the causes of people's thoughts, feeling and actions, coupled with a
complex comprehension of behavioural causality (please refer to Figure 1). At this level,
social phenomena can be logically and accurately explained. Psychological processes are
understood as impacting on actions, although causality is primarily understood as an
external force (Westen, 1985). Participants revealed an ability to describe internal
motivations as causal factors in their narratives.
4.2.7.1. TAT excerpt: Simple Internal Motivation
Participants often elaborated on psychological causes underpinning environmental or
behavioural causes. For instance, HB told the following narrative about Card 1:
"He tried to play the violin but he failed and now he doesn't try anymore. He sits and thinks
now first about his problems. He doesn't feel like going on. He's given up now. He feels
he's not good enough to play. He's just given up. He feels neglected; he feels bad. He tries
to break the violin or he tries to switch off everyone that's around him. He wants to be
alone. He doesn't want to communicate or talk to anyone."
In a similar vein, RB described a basic internal conflict on Card 6BM:
"He's thinking something otherwise. He's maybe going to steal someone's stuff or maybe go

and hurt someone. He's maybe going to hit. And he thinks now: he's going to do it now;
then he also doesn't want to."
55
4.2.7.1.1. Commentary on TAT excerpt: Simple Internal Motivation
Participants indicated an ability to elaborate on basic psychological tensions and causes,
reflecting an average awareness of internal processes. The cause of events was not singly
ascribed to external and visible behavioural causes, but acknowledged the presence of
unconscious processes involving a number of conflicting internal positions. Actions were
therefore not only understood as simple, unmotivated and reactive responses to external
forces, but as a result of internal and intentional motivations. Participants would thus seem
able to comprehend the internal antecedents of their violent behaviour, but are likely to be
overwhelmed by the malevolence and narcissism of the affective dimensions. Gross deficits
in the affective dimensions thus seem to underpin the enactment of violent behaviour.
4.2.8. Understanding of Social Causality: Individual Mean Scores
The overall mean score on social causality averaged Level 3, indicating a simple but logical
understanding of causal processes, although individual mean scores indicated minor logic
errors among half of the participants (please refer to Figure 5). Two of the participants
scored between Levels 3 and 4, reflective of a basic understanding of the role of
psychological events in motivating action. Six of the participants scored between Levels 2
and 3, indicating minor errors in logic. Examples of high (Levels 4 and 5) and low (Levels
1 and 2) scores will be given to illustrate the full range of responses.

56
40
Value Social causality
30
20
10
RP
RT
HB
RB
AM
RM
KF
MW
Respondent
Figure 5: Understanding of Social Causality
4.2.8.1. TAT excerpt: Complex Internal Causes Interspersed with Minor Logic Errors
A number of participants scored high level responses on this dimension. KF, in response to
Card 7BM, described a father and son acting on complex internal causes following a
murder charge and the boy's appearance in court:
"His father leans over to tell him: 'hey, don't worry you going home'. But in the meantime,
by the look of the boy, he knows that he's not going home. Maybe the father thinks it's for
the best for the son. The son has to learn by his mistakes and maybe he hasn't learnt yet.
That's why his father has to lie. But the son doesn't understand it. Deep down inside his
father's feeling bad; guilty. He knows it's wrong. He knows the consequences that his son is
gonna feel bad. But it's the best for him, it is for the best. The son's feeling very bad,
heartsore, cross, lot of anger. In the end the son finds out, well, what my father was doing
was right. At the end the boy changes his life. The father, he becomes proud of his son."

HB similarly demonstrated a complex understanding of the role of psychological events in
his response to Card 12M:
57
"Maybe the one that's now lying is busy getting a very bad dream. Cause maybe he's seeing
things that make him very fearful. Maybe he did something that worries him a lot. Maybe
killed someone. Or something like that. That's why he has bad dreams. His conscience
troubles him."
In stark contrast to the above, AM described actions as responses to simple feelings in his
narrative to Card 14:
"In his room are two girls that he's tied up; that he's misused. Then he dumps them. Cause
why? The women don't want to look at him cause at his house, in front of his house, are
very dangerous animals; dogs. He sets the dogs on people and that's why the people don't
want to look at him. They just think this man is a cruel man. And he then thought that if the
people think so, then I am so and he began to do so."
AM went on to illustrate an illogical causal process in an interaction with the character's
parents:
"Its not right that a man with such an age is without a wife. That's where crime comes
from."

4.2.8.1.1. Commentary on TAT excerpt: Complex Internal Causes Interspersed with
Minor Logic Errors
58
Much of the previous research on the understanding of causality among violent offenders
has highlighted disturbances in their cognitive processes. Sugarman (1980) described
idiosyncratic qualities in severely Conduct Disordered adolescents, Gacono & Meloy
(1994) identified fantasy-based relational attitudes among 100 Conduct Disordered
adolescents, while Gacono et al. (2000) noted deficits in reality testing and formal thought
disorder among psychopaths, sexual homicide perpetrators and paedophiles. Porcerelli et
al. (2001), in the only study using the TAT and SCORS, found a capacity to understand
events in causal and logical ways, interlaced with logic errors and unexplained transitions.
Participants' mean scores on causality indicated an understanding of simple psychologically
mediated causes, although half of the individual mean scores reflected minor logic errors.
These findings suggest that violent male juvenile offenders may at times demonstrate
illogical attributions and boundary confusion. The congealed deficits of a malevolent affect
tone and a narcissistic embeddedness are likely to find temporary satisfaction in the
projection of hated parts of the self through these illogical attributions and poor boundary
definitions. Moral injunctions are likely to be overwhelmed by these primitive and archaic
forces that seek sadistic control and the elimination of threatening external objects via
violent means.

59
4.3. CONCLUSION
Although a case study is not within the ambit of this research study, HB's mean scores on
the four dimensions will be used to conclude this chapter and illustrate one of the possible
manifestations of these internal patterns. HB, a seventeen year old youth, provided a
number of insightful responses in his TAT narratives, reflected in his complexity of
representations mean score midway between Levels 4 and 5. This high level of functioning
indicates an ability to make elaborate inferences about mental states, points of view,
motivations and unconscious processes. HB's mean score on social causality averaged
between Levels 3 and 4, indicating logical and accurate explanations of social phenomena
and a recognition of psychological factors in mediating action. Viewed to the exclusion of
these dimensions, HB would appear relatively insightful and conscious of the possible
motivations and consequences of both his and others' actions. Nevertheless, HB was found
guilty on a number of charges, including housebreaking, theft, robbery, armed robbery,
assault (x3) and murder.
HB's performance on affective dimensions does however offer insight into the perpetration
of these callous acts by an individual clearly capable of rational thought and complex
inference. HB averaged Level 2 on affect tone, indicating a pathologically dysphoric,
hostile, threatening and capricious view of the world. His mean emotional investment score
similarly averaged Level 2, reflecting a primitive need gratifying orientation with little
remorse following impulsive action. HB's pathological levels of functioning on these
affective dimensions reveals the potential for cognitively driven structures to be
overwhelmed and impaired in times of hyper-arousal, distress or object provocation. These

60
temporarily immobilised cognitive structures would seem to engender narcissistic
fulfilment through the mobilisation of malevolent affective components into gross
aggressive action. The inability to invest in others in meaningful ways may further serve to
protect HB from threatening intrusions into his fragile internal world, as well as to create a
lowering of empathy necessary to exert sadistic control over others. HB's object relational
patterns thus reveal the pivotal role played by the ascendancy of dysphoric and narcissistic
affective dimensions in the immobilisation of cognitively driven dimensions and the
enaction of gruesome acts among violent male juvenile offenders.

61
CHAPTER 5: RECOMMENDATIONS & CONCLUSION
5.1. Structural recommendations
The existence of only four Youth Care Centres nationally results in the majority of juvenile
offenders being located in sections of adult prisons (Blose, undated). The dearth of
rehabilitation programmes for juveniles within these facilities does not augur well for
attempts to address increasing levels of juvenile crime (DCS, 2005; Fagan, 2004). The
creation of greater numbers of juvenile correctional facilities, coupled with well-designed
rehabilitatory programmes, would seem essential in addressing a burgeoning problem. The
continued implementation of diversion programmes for juveniles found guilty of minor
offences would serve to facilitate a focus on serious offenders, while the provision of
statistics on juvenile crime categories would assist in delineating essential areas of focus.
5.2. Diagnostic recommendations
The majority of incarcerated juveniles are likely to fulfil criteria for both Oppositional
Defiant Disorder and Conduct Disorder (Dr. Koch, personal communication, 2004). As
previously discussed, the movement of some 25% of childhood Conduct Disorder
diagnoses to Antisocial Personality Disorder at eighteen years of age indicates the
persistence of juvenile personality characteristics into adulthood, although current
diagnostic criteria offer little insight into these underlying psychological attributes (Gacono
& Meloy, 1994; Moffit, 1993; Sadock & Sadock, 2003). Current American Psychiatric
Association (APA) diagnostic criteria for Oppositional Defiant Disorder and Conduct

62
Disorder are primarily behavioural means of categorisation that do little to ascertain
personality characteristics or individual variations among a highly heterogeneous grouping
of the population (American Psychiatric Association, 2000; Westen & Chang, 2000). These
diagnostic limitations impact negatively on the ability to distinguish serious character
problems from underlying personality conflicts, as well as the separation of adolescentlimited
conduct problems from life-course persistent antisocial behaviour (Moffit, 1993;
Gacono & Meloy, 1994).
The provision of revised diagnostic categories and criteria that encompass a wide range of
motivations underlying Conduct Disorder would seemingly be able to assist in the
identification of adolescents at risk for developing psychopathy later in life. Diagnostic
criteria could hypothetically outline a number of Conduct Disorder typologies, each with
differing internal constellations and projected outcomes. Irrespective of such possible
changes, projective tests will in all likelihood continue to play a vital role in the
identification of violent male juvenile offender object relations due to the intangibility of
personality constructs and the accompanying difficulty inherent in reducing complex
internal processes to simple present/absent categorisations that lack empirical validation
(Gacono & Meloy, 1994; Hare, Hart & Harpur, 1999; Westen & Shedler, 1999; Westen &
Chang 2000).
5.3. Treatment implications
Criminological explanations linking the genesis of violence to social deprivation appear to
be faltering (Legett, 2004). Attempts at diminishing violent male juvenile crime through the

63
amelioration of various forms of deprivation, including financial, social or familial, may
not have the desired effect if interpersonal expectations of others, which have been shown
to pervasively influence the current processing of social information, remain malevolently
skewed (Westen, 1985). A focus on sociological explanations and interventions to the
exclusion of personality characteristics contributing towards adolescent distress and
dysfunction is therefore not likely to address adequately the rising levels of violent male
juvenile crime.
Research findings among violent male juvenile offenders indicated higher levels of
functioning on cognitive dimensions relative to affective domains. The participants were
thus
able to infer basic psychological states, coupled with a logical and accurate
explanation of social phenomena through a simple psychological understanding of the
causes of thoughts, feelings and actions. A violent, aggressive, capricious and painful view
of the world accompanied a pathologically narcissistic orientation that offered limited
insight into the perpetration of callous acts. It seems reasonable to presume that relatively
sophisticated levels of cognitive functioning are overwhelmed by states of negative
affective arousal, leading to the violent enaction of a malevolent and need-gratifying
internal world.
Violent male juvenile offenders with similar object relational patterns to those identified
may cognitively comprehend the potential benefits of therapeutic interventions, although
the violent nature of their internal worlds may severely impair the development of a trusting
therapeutic relationship. In addition, therapists may be viewed with a narcissistic contempt
necessary to protect threatening intrusions into the offenders fragile internal worlds and so

64
permit their unempathic and sadistic control over others. Difficulty maintaining consistent
boundary definition, coupled with fluctuations in logical attributions may further contribute
to the projection of unwanted aspects of the self onto the therapist. Counter-transferences of
rage, or a wish to abandon their clients, may emerge in an obtuse attempt by the offenders
to confirm their pathological internal representations.
5.4. Research implications
Little research has been conducted on the underlying working representations of clinical or
non-clinical adolescent samples (Westen & Chang, 2000). Violent male juvenile offenders
appear to represent an essential area of research given the escalation of violent crime in this
population grouping, as well as the extreme and unremitting frequency of crime committed
by a small number of psychopathic males (DCS, 2005; Moffitt, 1993; Gacono & Meloy,
1994; Sadock & Sadock, 2003). Research involving a larger number of violent male
juvenile offenders, as well as a control group, would permit inter-group and crossdimensional
statistical comparisons, as well as generalizability. Given sufficient resources,
further research may wish to explore object relational patterns of different types of violent
male juvenile offenders. Comparative research between violent male juvenile and adult
offenders may offer valuable insight into the continuation of adolescent personality
pathology into adulthood.
The use of TAT and SCORS in assessing object relational patterns may not however
represent the most efficient tool to do such research. The extensive and labour-intensive
training required by the coding manual, as well as the intricate scoring procedure, result in

65
an inordinate amount of time required for data analysis that may not be possible within the
constraints of psychological services offered in correctional facilities. Thematic analysis of
TAT responses or the creation of a computerised version of SCORS may represent viable
alternatives. Failing this, less labour intensive projective techniques may need to be utilised.
5.5. Conclusion
In this study, object relational patterns were ascertained through the analysis of TAT
narratives using SCORS. Overall, findings indicated lowered scores on capacity for
emotional investment and affect tone relative to complexity of representations and
understanding of causality. Participants thus demonstrated elevations on the cognitive
dimensions relative to the emotional domains. These findings suggest that violent male
juvenile offenders are able to infer basic psychological states, as well as logically and
accurately explain social phenomena through a simple psychological understanding of the
causes of thoughts, feelings and actions. Viewed in isolation, the results on the cognitive
dimensions may well be presumed to indicate an average level of general functioning.
Performances on the affective dimensions did however offer some insight into the callous
acts perpetrated by the participants. The juvenile offenders demonstrated a violent,
aggressive, capricious and painful view of the world entailing little expectation of
interpersonal co-operation. Participants also revealed an accentuated need gratifying
orientation and absence of remorse in the face of violent action. The specific combination
of these object relation patterns may in all likelihood contribute towards violent action in
particularly emotionally distressing moments. While folklore may imagine psychopaths as

66
murdering, raping and assaulting twenty four hours a day seven days a week, results
indicate malevolent and narcissistic imperatives to be mediated largely by average levels of
cognitive functioning. It seems that certain overwhelmingly distressful circumstances are
likely to activate pathological emotional dimensions, resulting in a loss of perspective
normally offered by the cognitive realms.
In addition, and consistent with previous Rorschach studies of sexual homicide perpetrators
(Gacono et al., 2000; Meloy et al., 1994) and Porcerelli et al's. (2001) case study of a
psychopathic serial sexual homicide perpetrator, variations in scores were only found across
cognitive dimensions. Individual mean scores on complexity of representations indicated an
ability to represent others multidimensionally while simultaneously struggling with the
definition of characters in clearly bounded ways. Similarly, results on the social causality
dimension indicated a number of illogical attributions. These variations, coupled with a
consistently dysphoric affect tone and narcissistic orientation to the world, may in all
likelihood facilitate the projection of unwelcome and tension-inducing aspect of the self,
thereby eliciting responses of rage or abandonment and so confirming simple and
malevolent internal representations. The confirmation of these hostile representations may
serve to further enhance malevolent and narcissistic imperatives, thereby necessitating the
destruction of perceived external threats in a perverse attempt to restore a delicate internal
equilibrium. Certain situational contexts thus appear able to overwhelm the violent male
juvenile offenders' ability to comprehend the multidimensional nature of others, as well as
the basic tenets of social causality, resulting in the violent enaction of a grossly malevolent
and narcissistic internal world.

67
REFERENCES
American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental
Disorders (4 th ed Text Revised). Washington: American Psychiatric
Association.
Appelbaum, S. & Mayam, M. (1974). A Rorschach test system for understanding
personality. Unpublished manuscript, the Menninger Foundation, Topeka, KS.
Article 40, (2000, February). Hard cases... Available at: www.communitylawcentre.org.za,
2(1).
Atkinson, R., Atkinson, R., Smith, E., Bem, J. & Hilgard, E. (1985). Introduction to
Psychology. Tenth Edition.
Barday, R. (2001). Sexual assault cases in the news. Gender News, (5),1.
Bellak, L., Pasquarelli, B.A., & Branerman, S. (1949). The use of the Thematic
Apperception Test in psychotherapy. Journal of Nervous and Mental Diseases,
(110), 51-65.
Bernstein, D., Cohen, P., Velez, C., Schwab-Stone, M., Siever, L. & Shinsato, L. (1993).
Prevalence and stability of DSM-III-R personality disorders in a communitybased
survey of adolescents. American Journal of Psychiatry, 153, 907-913.
Blatt, S. & Lerner, H. (1983). Investigation in the psychoanalytic theory of object relations
and object representations in J. Masling (Ed), Empirical Studies in
Psychoanalytic Theories Vol. 1. Hillsdale: Lawrence Erlbaum Associates.
Blose, T. (undated). Report prepared for the United Nations Department of Justice on the
rights of the child. Unpublished report.
Cape Argus (2005, January 18). Beaten to death by classmate.
Chisholm, L. & September, J. (2005). Gender equity in South African education. 1994-
2004: conference proceedings. Cape Town: HSRC press.
Cowden, J., Bassett, H., Cohen, M. (1969). An analysis of some relationships between
fantasy-aggressive and aggressive behaviour among institutionalized
delinquents. The Journal of Genetic Psychology, 114, 179-183.
DCS. (2001). Department of Correctional Services Annual Report 1 January 2000
to 31 March 2001. Cape Town: Formeset.
DCS. (2002). Department of Correctional Services Annual Report 1 April 2001 to 31
March 2002. Cape Town: Formeset.

DCS (2003). Department of Correctional Services Annual Report 1 April 2002 to 31
March 2003. Cape Town: Formeset.
DCS (2004). Department of Correctional Services Annual Report for the 2003/04
Financial Year. Cape Town: Formeset.
DCS (2005). Department of Correctional Services Annual Report for the 2004/05
Financial Year. Cape Town: Formeset.
De Wet, N. (2005). Bullying in schools. Everyone's problem. Unpublished inaguaral
lecture (1 June 2005). Bloemfontein: University of the Free State.
Die Burger (2004, May 6). Nie net arm kinders raak betrokke by gewelddadige oortredings
nie, se Nimro.
Die Burger (2004, May 26). Moord en verkragting in Khayelitsha kwel staat.
Die Burger (2004, May 28). Kaapstad het meeste moorde.
Die Burger (2004, June 16). DKD wil jeugdiges rehabiliteer.
Die Burger (2004, June 4). Jeug mishandel in oorvol SA tronke.
Die Burger (2004, July 10). Staat se of jong kind vervolg kan word.
Die Burger (2004, August 24). Skolier van Langa ook in klas gesteek.
Die Burger (2004, October 7). Al meer geweld in skole.
Die Burger. (2005, February 28). Vergaderings kom oor vlaag van geweld by Kaapse skole.
Die Burger (2005, March 18). Rasool gryp in oor geweld by skole.
Die Burger (2005, April 11). Skolegeweld is ergste in Wes-Kaap.
Die Burger (2005, April 12). Kruispad 'n kookpot van misdaad en verval – studie.
Die Burger (2005, April 13). Skoolhoof getuig nadat leerder (7) verkrag is.
Die Burger (2005, April 13). Byna 100 000 wapens 'verloor'.
Die Burger (2005, April 14). Balfour wil gevangenes 'wegwys'.
Die Burger (2005, April 26). 'Byna geen hulp vir jong seksoortreder'.
Die Burger (2005, June 15). Waterkloof-vier is skuldig aan moord.
68

Die Burger (2005, June 15). Meisie 'kom te lig daarvan af' nadat sy onderwyser aanrand.
Die Burger (2005, July 2). Nagenoeg 18 000 al uit tronke vrygelaat.
Die Burger (2005, July 11). Baba Jordan se dood tref bewoe ma nou eers.
Die Burger (2005, July 22). Iets moet nou gedoen word aan die geweld in Manenberg.
Die Burger (2005, August 8). Nuwe wet 'sal onderwysers kan beveilig'.
Die Burger (2005, August 9). Skok van aanval in klas bly onderwyseres steeds by.
Die Burger (2005, September 5). Dertienjarige voor vir moord.
Die Burger (2005, September 16). Regering dalk hof toe oor toestande in SA tronke.
Die Burger (2005, September 20). Baden Powell-verdagtes vas oor 'moord-poging'.
Die Burger (2005, October 13). Twee tieners in hof na seun (9) se skietdood.
Dodge, K., Murphy, R. & Buchsbaum, K. (1984). Social cognitive biases and deficits in
aggressive boys. Child Development, 52, 389-409.
Dodge, K. & Somberg, D. (1987). Hostile Attributional Biases among Aggressive Boys Are
Exacerbated Under Conditions of Threats to the Self. Child Development, 58,
213-224.
Du Plessis, A. & Louw, A. (2005). The tide is turning. The 2003/2004 SAPS crime stats.
Institute for Security Studies. SA Crime Quarterly. No. 12.
Fagan, H. (2004). Laws on minimum sentencing and parole spell worsening prison
conditions. Institute for Security Studies. Crime Quarterly. No. 10.
Feshback, S. (1955). The Drive-Reducing Function of Fantasy Behaviour. Journal of
abnormal and social psychology, 50, 3-11.
Gacono, C., Meloy, R. & Heaven, T. (1990). A Rorschach Investigation of Narcissism and
Hysteria in Antisocial Personality. Journal of Personality Assessment, 55 (1),
270-280.
Gacono, C. & Meloy, J. (1992). A Psychotic (Sexual) Psychopath: "I just had a violent
thought ...". Journal of Personality Assessment, 58(3), 480-493.
Gacono, C. & Meloy, J. (1993). A Borderline Psychopath: "I was basically maladjusted ...".
Journal of Personality Assessment, 61(2), 358-373.
69

70
Gacono, C. & Meloy, J. (1994). The Rorschach Assessment of Aggressive and
Psychopathic Personalities. Hove: Lawrence Erlbaum Associates.
Gacono, C., Meloy, J. & Bridges, M. (2000). Sexual Homicide Perpetrators, and
Nonviolent Paedophiles. Journal of Clinical Psychology, 56 (6), 757-777.
Hare, R., Hart, S. & Harpur, T. (1999). Psychopathy and the DSM-1V Criteria for
Antisocial Personality Disorder. Journal of Abnormal Psychology, 100, 391-
398.
Hare, R.D., Harpur, T.J., Hakstian,. A.R., Forth, A.E., Hart, S.D. (1990). The Revised
Psychopathy Checklist: Reliability and Factor Structure. Psychological
Assessment, 2, 338-341.
Harter, S. (1988). The development of self-representation in N. Eisenberg (Ed), Handbook
of Child Psychology: Vol 3. Social, Emotional and Personality Development.
New York: Wiley.
HSRC Annual Report (2005). South African Public Schools. Cape Town. HSRC Press.
James, P., & Mosher, D. (1967). Thematic Aggression, Hostility-Guilt and Aggressive
Behaviour. Journal of projective techniques and personality
assessment, 31, 61-67.
Jenkins-Hall, K., & Osborn, C. (1994). The conduct of socially sensitive research: Sex
offenders as participants. Criminal Justice and Behaviour, 21(3), 325-340.
Kagan, J. (1956). The measurement of overt aggression from fantasy. Journal of abnormal
and social psychology, 52, 390-393.
Kagan, J. (1959). The stability of TAT fantasy and stimulus ambiguity. Journal of
Consulting Psychology, 23, 266-271.
Keltikangas-Jarvinen, L. (1982). Alexithymia in Violent Offenders. Journal of Personality
Assessment, 46(5), 462-467.
Kernberg, O. (1975). Borderline Conditions and Pathological Narcissism. New York: Jason
Asonson.
Legett, T. (2003). The Fact behind the Figures. Crime Statistics 2002/2003. SA Crime
Quarterly, 6.
Legett, T. (2004a). What's Up In The Cape? Crime rates in Western and Northern Cape
Provinces. SA Crime Quarterly, 7.
Legett, T. (2004b). Still Marginal. Crime in the coloured community. SA Crime Quarterly,
(7).

Lesser, G. (1957). The relationship between overt and fantasy aggression as a function of
maternal response to aggression. Journal of abormal social Psychology, 52,
390-393.
Lesser, G. (1958). Conflict Analysis of fantasy aggression. Journal of Personality, 26, 29-
41.
Lewinsohn, P., Rohde, P., Seeley, J., Klein, D. (1997). Axis 11 Psychopathology as a
Function of Axis 1 Disorders in Childhood and Adolescence. Journal of
American Academic Child & Adolescent Psychiatry, 36(12), 1752-1759.
Masling, J. (1997). On the Nature and Utility of Projective Tests and Objective Tests.
Journal of Personality Assessment, 69(2), 257-270.
McManus, M., Alessi, N., Grapentine, W. & Brickman, A. (1984). Psychiatric Disturbances
in Serious Delinquents. American Journal of Child and Adolescent Psychiatry,
23, 602-615.
Mistry, D. (2004). Falling Crime, rising fear. Crime quarterly. Institute of Security Studies.
Volume 8.
Moffitt, T. (1993). Adolescence-Limited and Life-Course-Persistent Antisocial Behaviour:
A Developmental Taxonomy. Psychological Review, 100, 674-701.
Muntingh, L.M. (2001). Update: Sentenced and unsentenced children in prisons in Article
40, 3(1). Available at: www.communitylawcentre.org.za
Muntingh, L. M. (2003). Children in Conflict with the Law: A Compendium of Child
Justice Statistics. Cape Town: Nicro.
Murray, H. (1943). Thematic Apperception Test manual. Cambridge: Harvard University
Press.
Mussen, P. & Naylor, H. (1954). The relationship between overt and fantasy aggression.
Journal of abnormal social Psychology, 49, 235-240.
Myers, W., Barket, R. & Harris, E. (1995). Adolescent Psychopathy in Relation to
Delinquent Behaviours, Conduct Disorder, and Personality Disorders. Journal
of Forensic Sciences, 40, 435-439.
Nasby, W., Hayden, B., & de Paulo, B. (1979). Attributional bias among aggressive boys to
interpret unambiguous social stimuli as displays of hostility. Journal of
Abnormal Psychology, 89, 459-468.
Ornduff, S., Freedenfeld, R., Kelsey, R., & Critelli, J. (1994). Object Relations of Sexually
Abused Female Subjects: A TAT Analysis. Journal of Personality Assessment,
63(2), 223-238.
71

Osborne, C., (1996). The Feasibility & Evaluation of the Thematic Apperception Test for
adolescent clients. Measurement & Evaluation in Counselling & Development,
29, 48–56.
Porcerelli, J., Abramsky, M., Hibbard, S. & Kamoo, R. (2001). Object Relations and
Defense Mechanisms of a Psychopathic Serial Sexual Homicide Perpetrator: A
TAT Analysis. Journal of Personality Assessment, 77, 87-105.
Purcell, K. (1956). The TAT and anti-social behaviour. Journal of consulting Psychology,
20, 449-456.
Rain, A., O' Brien, M., Smiley, N., Scerbo, A. & Chan, C. (1990). Reduced Lateralization
in Verbal Dichotic Listening in Adolescent Psychopaths. Journal of
Abnormal Child Psychology, 99, 451-463.
Rule, B., Nesdale, A. & McAra, M. (1974). Children's reactions to information about the
intentions underlying an aggressive act. Child Development, 45, 794-798.
Rustomjee, R. (2005). Public lecture at the University of the Western Cape's Institute for
Infectious Illness and Molecular Medicine.
Sadock, B. & Sadock, V. (2003). Kaplan & Sadock's Synopsis of Psychiatry. 9 th Edition.
USA: Lippincott Williams & Wilkins.
Sanford, R., Adkins, M., Miller, R., Cobb, E. (1943). Physique, personality and
scholarship: a cooperative study of school children. Monogr. Soc. Res. Child
Developm., 8.
Schaefer, J. & Norman, M. (1967). Punishment and aggression in fantasy responses of boys
with antisocial character traits. Journal of Personality and Social Psychology, 6
(2), 237-240.
Sekhonyane, M. (2004). First Things First. Rehabilitation starts with alternatives to prison.
SA Crime Quarterly, 7.
Shore, M., Massimo, J., & Mack, R. (1964). The relationship between levels of guilt in
thematic stories and unsocialised behaviour. Journal of Projective Techniques,
28, 346-349.
Sugarman, A. (1980). The borderline personality organization as manifested on
psychological tests in J. Kwawer, P. Lerner, H. Lerner, & A. Sugarman.
(Eds.), Borderline phenomena and the Rorschach test. New York:
International Universities Press.
Sugarman, A. (2000). The Contribution of Psychoanalytic Theory to Psychological Testing.
Psychoanalytic Psychology, 17 (1), 3-23.
72

Thomson, J. (2004). A Murderous Legacy. Coloured homicide trends in South Africa. SA
Crime Quarterly, 7.
Westen, D. (1985). Social Cognition and Object Relations Scale (SCORS): Manual for
Coding TAT Data. Unpublished Manuscript.
Westen, D. (1989). Are “Primitive” Object Relations Really Preoedipal? American
Orthopsychiatric Association, 59, 331-345.
Westen, D., Ludolph, P., Lerner, H., Ruffins, S., Wiss, C. (1990a). Object Relations in
Borderline Adolescents. American Academy of Child and Adolescent
Psychiatry, 29, 338-348.
Westen, D., Ludolph, P., Silk, K., Kellam, A., Gold, L., & Lohr, N.E. (1990b). Object
Relations in Borderline Adolescent and Adults: Developmental Differences.
Adolescent Psychiatry, 17, 360-384.
Westen, D., Lohr, N., Silk, K., Gold, L. & Kerber, K. (1990c). Object relations and social
cognition in borderlines, major depressives, and normals: A thematic
apperception test analysis. Psychological Assessment: A Journal of Consulting
and Clinical Psychology, 2, 355-364.
Westen, D. & Arbor, A. (1990d). Towards a revised theory of borderline object relations:
contributions of empirical research. International Journal of Psycho-Analysis,
71, 661-693.
Westen, D., Klepser, J., Ruffins, S., Silverman, M., Lifton, N., & Boekamp, J. (1991a).
Object Relations in Childhood and Adolescence: The Development of
Working Representations. Journal of Consulting Clinical Psychology, 59, 400-
409.
Westen, D. (1991b). Clinical Assessment of Object Relations Using the TAT. Journal of
Personality Assessment, 56(1), 56-74.
Westen, D. (1991c). Social Cognition & Object Relations. Psychological Bulletin, 109,
429-455.
Westen, D. & Shedler, J. (1999). Revising and Assessing Axis11, Part 1: Developing a
Clinically and Empirically Valid Assessment Method. American Journal of
Psychiatry, 156 (2) 258-272.
Westen, D. & Chang, C. (2000). Personality Pathology in Adolescence: A Review.
Adolescent Psychiatry, 25, 61-100.
Wesley, M. (2002). Origin and History of the Earliest Thematic Apperception Test Pictures.
Journal of Personality Assessment, 79, 422-446.
73

Wood, C. (1998). A profile of young sex offenders in South Africa: A Pilot Study.
Unpublished thesis. Rhodes University.
74

75
APPENDIX A: A Brief Synopsis of Measures of Object Relations
(From: Westen, 1991b: 56-74)